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

S&P make the news… with a Trend Following Index

August 10th, 2011 · 34 Comments · Equities, Fund Review, Futures, Trend Following

They say any publicity is good publicity

If that’s the case, Standard and Poor’s might have achieved the “marketing coup” of the year/decade/century (time will tell) with their downgrade of US credit rating last Friday (possibly thanks to a $2 Trillion Mistake).

SGMI: a new Trend Following Index

But this is not the piece of news that caught my attention today.
Standard and Poor’s announced the creation of the S&P Systematic Global Macro Index (SGMI).

With this index, S&P basically intends to track – or rather replicate – the performance from the Managed Futures / CTA space:

London, August 9, 2011 – S&P Indices has launched the S&P Systematic Global Macro Index (SGMI), which aims to reflect price trends of highly liquid global futures, representing the general level of volatility taken by managers in the global macro and managed futures/Commodity Trading Advisor (CTA) space.

This is a rules-based index, which will implement a trend following system to trade a (semi-diversified) portfolio of liquid futures markets, with a fairly standard risk-adjusted equal allocation per sector and per instrument:

The Index is diversified globally across 37 constituents, falling into the six most widely traded sectors– Commodities, Energy, Fixed Income, Foreign Exchange, Short Term Interest Rates and Equity Indices.

The weighting scheme applies an even risk capital allocation across the index by sector and again to each constituent within each sector so that no single sector or constituent drives the volatility of the index.

Other Trend Following Indices

Of course, this is not a new concept. There are several companies tracking the performance of Managed Futures (Newedge, BarclayHedge) with their respective indices – or even “closer to home” the Trend Following Wizards report on this blog.

Even the concept of using a rules-based systematic trend following approach has been implemented before. Apart from the State of Trend Following monthly report, also on this blog, long-time readers will recall this Beta of Managed Futures paper by Conquest Capital Group, which developed a mechanical trend following benchmark aimed at replicating CTAs performance. I discussed the paper more than a year ago in this “Betafication of Alpha: towards a Commoditization of Trend Following?” post, and was asking then “when the launch of a Trend Following ETF?“.

The end game of Conquest is of course to market their fund implementing this benchmark and it is pretty certain that the end game with this S&P index is similar. One can only speculate (again) as to how long it will take to see a SGMI Trend Following ETF.

The underlying strategy in this new S&P index seems to contain an adaptive timeframe logic (whereas the Conquest Capital Group benchmark uses a combination of several fixed timeframes to capture trends at different levels):

Uniquely, the trend-following model used [by the SGMI] to determine the position of each constituent is flexible enough to allow a customized time-period for each constituent on a monthly basis, unlike models which are based on fixed time-periods. This means that if a longer term trend is driving the market, the Index reflects that, but if a shorter-term trend becomes significant the Index picks that up, using an iterative process to test the stability of each trend.

Trend Following ETFs: a Viable Possibility?

Of course, despite the potential appeal of strong replication and much lower fees for a “speculative” SGMI ETF, this is not necessarily a winning proposition for the investor. The debate resides in whether CTA managers can justify their fees.

Using a Geometric Information Ratio calculation based on the Conquest benchmark, I showed previously that CTAs still produce alpha over the benchmark on a net-of-fees basis – most likely thanks to their research in other areas such as enhanced rolling methodology, execution efficiency, extra diversification (100+ markets instead of 37), etc.
Note that this calculation was not free of survivorship bias though.

This could be one of the reasons why the Conquest fund has not taken off in any substantial manner (less than $200M in AUM after 7 years trading).

Another similar Trend Following fund is the Cambria Global Tactical ETF (GTAA), run by Mebane Faber. Unlike CTAs it does not trade futures (it is an actively managed “ETF of ETFs”), however its core strategy uses long-term trend following principles to allocate funds between 50 to 100 underlying ETFs representing different global asset classes.

It only launched late last year and is already at the same levels of AUM as the Conquest fund, showing that there is an appetite for this type of product in an accessible format (i.e. no high minimum account size).

A fact also highlighted in the S&P press release:

“Issues like high minimums and high fees have made it difficult for many investors to gain access to global macro and managed futures strategies. We envisage that new products based on this index will give investors the ability to invest in a long/short, comprehensive set of the main futures contracts.”

“Gimmicky” Index and ETF?

Josh Brown and Abnormal Returns argue that, in a case of the tail wagging the dog giving birth to puppies, there is a plethora of index creation, in the sole aim to support “niche and gimmicky ETFs” – quickly turning into “zombie ETFs” destined to populate the ETF Deathwatch list.

It will be interesting to see if the “SGMI ETF” is one these “puppies” or whether it can create a substantial impact in the Managed Futures and Trend Following industry – as its great-grandfather the SPY did within the wider investment industry (trading volume: $80B yesterday).

What To Think of it?

It obviously depends on the actual performance of this index and its derived products, so only time will tell.

CTAs: Do you see this development as a threat to your business, or as a way to bring more awareness to the sector? Are you feeling confident about the added value that you can provide to fend-off this potential new competition?

Investors: Are you excited at the prospects of being able to invest in managed Trend Following without the big “minimum account size” hurdle? Are you skeptical of the capacity of the index to replicate the performance of Trend Followers?

Independent system developers: Does this reinforce your belief into building your own system? Would you reconsider building a system if you could invest in an ETF-like Trend Following product?

I’m keen to hear your opinions… Feel free to contribute to the discussion in the comments section below.
UPDATE: Tim Pickering from Auspice Capital has commented below but also expressed his (interesting) point of view from a CTA perspective on their blog. In essence they do not feel threatened as they “believe in the separation of the alpha and index methods” (read the post here to see why). He also pointed out that:

S&P is surely not the first to do this. In fact, their previous effort has been widely regarded as unsuccessful from a performance standpoint. The S&P DTI (and related) indices have been around for some time now.

UPDATE 2: Several Trend Following investment products (ETF, ETN, etc.) already exist actually.

Prompted by reader Pumpernickel’s comment (below), I decided to research more about the DTI index that was mentioned by Tim Pickering in his blog post. I was under the impression that a trend following ETF was a really new concept, but it appears that some products very similar already exist. Below is a summary from this “google” research:

The “Diversified Trends Indicator” (DTI) has been created by Victor Sperandeo, with a few investment products having been launched based on this index (or one of its sub-indices: the “Commodity Trends Indicator” [CTI] and the “Financial Trends Indicators” [FTI]). The DTI is a based on a single mechanical trend following/momentum strategy applied to a portfolio of 24 US futures.

From Alpha Financial Technologies (Victor Sperandeo’s firm) website:

In 2002, AFT granted Standard & Poor’s the exclusive right to sublicense the indexes to third parties, known as the S&P Diversified Trends Indicator, S&P Commodity Trends Indicator, and S&P Financial Trends Indicator. S&P launched the S&P Diversified Trends Indicator in January of 2004. Over the next two years several offshore products linked to the S&P DTI were launched, including a UCITS III Fund by Nomura International PLC. In 2007, Rydex launched the first long/short managed futures mutual fund, which tracks the S&P DTI. Direxion funds launched a mutual fund that tracks the CTI® and a mutual fund that tracks the FTI™ in 2008 and 2009, respectively. In July of 2008, Merrill Lynch structured an exchange-traded note that tracks the S&P CTI.

In November 2009, AFT commenced licensing its indexes directly to third parties, although existing S&P licenses remain in effect. On August 1, 2010 AFT launched the FX Trends Index™ (FXTI®). In January 2011, Wisdom Tree launched an exchange traded fund (ETF) which tracks the DTI®.

Investment banks and financial institutions in over 15 countries worldwide have licensed AFT’s indexes. As of April 2011, there are over $3 billion invested globally in products utilizing AFT’s indexes.

Tickers for the indices and some of their investable products:

Bloomberg Index Tickers:

  • S&P Diversified Trends Indicator: Price Return: SPDTP, Total Return” SPDTT.

Investable Products:

  • Rydex SGI Managed Futures: RYMFX (mutual fund based on S&P DTI and trading since 2007)
  • ELEMENTS S&P CTI: LSC (ETN based on the S&P CTI and trading since 2009)
  • Direxion Commodity Trends: DXCTX (mutual fund based on the AFT CTI and trading since 209)
  • Direxion Financial Trends: DXFTX (mutual fund based on the AFT FTI and trading since 209)
  • WisdomTree Managed Futures: WDTI (ETF based on the AFT DTI and trading since 2011)

The methodology used to calculate the index is made public by AFT – it can be found here as a pdf document.
A document for the S&P DTI can be found there as a pdf document.

I have also found this “interesting discussion” between Attain Capital and Victor Sperandeo on whether DTI-based products really can deliver Managed Futures performance.

In the first “missive” (in this article), Attain Capital argue that the DTI-based products are wrongly labeled as “Managed Futures” as they offer no exposure to the sector directly. They go on to show that the DTI index has seriously under-performed the Newedge CTA Index (and that in turn the Rydex mutual fund slightly underperformed its DTI benchmark due to expenses) since 2007 (total return of 28.85% for the Newedege index vs. 3.36% for the Rydex fund). They also chart the 12-month rolling correlation between the DTI and various CTA indexes. The figure oscillates around 0.3 and 0.9 for an average of 0.6 between 2000 and 2007.

Victor Sperandeo responds that he “developed the Diversified Trends Indicator™ as a way for investors to access the “core” returns embedded within trend-following in the futures markets” and that the DTI does indeed implement a (single) Managed Futures strategy by the way of a mechanical trend following system on a diversified portfolio of futures. He further highlights that the DTI actually out-performed the Newedge CTA Index since 2004 while providing a more practical way of investing in Managed Futures, compared to the un-investable CTA index from Newedge.

In a further reply, Attain Capital highlights that comparison between DTI and Newedge CTA Index should be done on a realistic basis (DTI Total return minus an annual 2% fee with the Newedge index already being net of fees), in which case the DTI slightly under-performs the Newedge CTA index, concluding:

  We don’t believe they will see long term success beating the managed futures indices (or if they are even trying to beat those benchmarks) because they track only a single strategy.

It seems that both have a point – and certainly a different viewpoint (and interests). It actually ties back to the alpha vs. replicated beta “debate” and what place the mechanical and investable index-based products take in the managed futures space.

As Tim Pickering was saying “I believe in the separation of the alpha and index methods. We feel it will bring awareness to CTA and offer product to investors that do not know about/understand or have access to “Accredited or QEP” products. For CTAs that generate Alpha at reasonable price, this exposure can only help you.”

It will be interesting to see – but hard to measure – whether a Starbuck effect additionally develops from these products to boost sales from classic CTAs.

These indices and products are something I had completely missed, so thanks again to the readers who offered pointers to more info.
UPDATE 3: Trader Vic Index (TVI):

Thanks to reader RB, who pointed out the Arrow Managed Futures Trend Fund, which tracks another similar index created by Victor Sperandeo: the Trader Vic Index (TVI).

The index was developed as a partnership between RBS and Enhanced Alpha Management, L.P. (EAM, LP) – Sperandeo’s CTA firm – and launched in 2009. The Arrow Funds mutual fund offering based on that index started trading in 2010, and has around $100M in AUM.

Tickers: TVICTR:IND (Bloomberg index ticker) and MFTFX (Arrow mutual fund on google finance).

Related Posts with Thumbnails


34 Comments so far ↓

  • Wednesday links: indifferent markets | Abnormal Returns

    [...] Where are all the managed futures ETPs?  (World Beta, Au.Tra.Sy Blog) [...]

  • Reformed Trader

    Thanks for the helpful information, Jez.

    “Are you skeptical of the capacity of the index to replicate the performance of Trend Followers?”

    Yes. If other trend followers are like me, they’ve developed systems that scale into and out of positions. Even if we didn’t do that, we’re small enough that no one will notice us. I’m concerned that funds that mechanically follow binary signals will be an easy target and suffer the same fate as USO.

    “Does this reinforce your belief into building your own system? Would you reconsider building a system if you could invest in an ETF-like Trend Following product?”

    I would build the system anyway, mostly because trend-following is not the only strategy I use. It’s more capital efficient to build trend-following into a portfolio using swarm management than to try to buy a separate ETF for each strategy.

  • RB

    I don’t think the CTAs will feel threaten by this. Anyone that has that much money will still put money with the best CTAs. It would be great for the small guy. He could include it as part of a asset allocation model. You could combine this with a value fund manager . I once combined Bill Dunn’s returns with Jean Marie Eveillard’s returns and it smooth out the annual returns. System developers should still build their own systems and if a good trend following product comes out. Then you could also use it to diversify with your system. I must say that this site is one of the most sophisticated and knowledgeable sites out there, and it’s comment section has no equal. The people on here are of the highest caliber.

  • Jez Liberty

    @Reformed Trader: good point re: the analogy with USO. Thanks for the comment.

    @RB: Thanks for the kind words!
    I tend to think along the lines of your thinking (and similarly to Reformed Trader’s point of view). CTAs will probably still have the upper hand in terms of performance (and AUM from HNWI or institutions) thanks to the added value coming from on-going research – but a TF ETF could be a good enough proxy to tempt the retail investors who are currently barred from investing with CTAs because of high minimum accounts.

  • HumbleSystemBuilder

    In Canada Claymore launched a managed futures ETF (Auspice Capital designed the index). The ETF closed due to lack of investment but the long only version is still open.
    Maybe Canada was not a deep enough market?

    With regards to @reformed USO comment, that is a good point I agree it could be a concern, there are algo’s out there designed to find or guess where a bulk of orders are and take advantage of this. The ETF would have to be quite large though for this to happen.

    Furthermore USO has changed there roll execution to have reduce this concern. I have never reverse engineered USO but I would guess that long-term investing (not trading) in a “contango” commodity is what hurts the ETF the most of “truly” hurts them.

    Regarding research, the index is likely going to use stationary parameter(s), CTA’s like to use walk forward optimization and other adapting techniques to tweak their systems over time. Currently we are experience a very difficult short-term system trading environment over the past 2 years, strong CTA’s are adapting to this and changing accordingly. Furthermore top CTA’s use risk management overlay’s that help reduce drawdowns.

    Lastly, top CTA’s have top trading teams, the ETF would not have this. I think anyone who has the time to think this much about asset allocation has enough money to invest in real CTA’s. If you dont have enough money you are probably real interested in building your own system and wont care about the ETF.

    System trading is a highly specific discipline, the best index is an index of real CTA performance.

    We dont need to compare the CTA’s to a silly benchmark like this, we compare CTA’s to other CTA’s and look at their relative returns and risk management, that is how we evaluate CTA’s.

    Evaluating equity managers against the index and evaluating CTA’s against an index is not apples to apples. This is why a system trading ETF will never be able to what the SPY is to equities. Evaluating CTA’s against the benchmark CTA index is the correct substitute. Creating a barclay’s or stark CTA index ETF would be very difficult if not impossible. This would be the acceptable alternative.

    Great site!

  • Jez Liberty

    Hi HumbleSystemBuilder (I like that moniker!).
    Thanks for your thoughts and comment on this. I was not aware of the Claymore ETF in Canada, but this is interesting to know that it failed to take off.
    You make very good points and these are probably a reason why there is a discrepancy between the claims in the Conquest paper and their actual live performance compared to CTAs.
    I believe it will still be interesting to see what the retail investor demand is for this product (Meb Faber’s ETF seems to be doing alright for now for example).
    Thanks again.

  • ZigZag

    Thanks for all the effort and time you put into this blog – the output is excellent.
    I have had capital at risk on a long-term trend-following system for about 1 year now. So far, so good, even though it is going through a deep DD at this point. I also study and develop systems as a (serious) hobby.
    As a consumer of rules-based investment management products, I would put any TF ETF to very high standards of units of cost per units of risk-adjusted alpha and of track-record – the same standards that I would require of a CTA. Further, I value disclosure and the ability to control and test the logic and parameters of a system. The availability of a TF ETF would definitely not take away from my system-developing efforts. But that’s a mainly a personal decision, as my approach to trading my own capital is do-it-yourself, for good or bad. Having said that, a TF ETF is an interesting product and I would probably risk some capital on it, if nothing else to have it on my radar screen, so to speak. It should also be a more affordable way to invest in trend-following, albeit, I imagine, with some (acceptable) tracking error relative to the true performance of CTAs (or of some TF benchmark strategy? Which one? More to this pont: if a TF ETF will track CTAs then by definition this is a self-defeating idea at the limit, as the success of the TF ETF will detract from CTAs business which would in turn make CTAs an inviable benchmark…Did I miss the point?). In any case, a TF ETF does not strike me as product for mass-consumption but rather for a niche market.

  • Praetorian

    I have a couple of comments about this:

    1. Most ETFs are designed by people who think like equity traders/investors, so I don’t think they will be able to build anything of worth. Actually, I was saw an explanation by a famous trader of the rules applied to his Commodities Index Fund, I was appalled by the lack of sophistication.

    2. If you have an ETF, you have to publish the rules, that means people can trader ahead of you, I met a CTA in Chicago who made a ton of money trading ahead the long only commodities funds before they had to roll, the easy money didn’t last because that was too easy.

    3. I would be willing to try an Expert Trend Follower is behind developing the ETFs, but I don’t think that is likely.

    4. But most importantly, there is a CTA FOF in Ireland called Abbey Capital, they are invested in the best trend followers, have super low minimums (like 1k EUR) and daily liquidity, this is probably the best option for the small guy.

  • Jez Liberty

    @ZigZag: Thanks for your comment and kind words. Agree on the fact that this would still be “niche”. Managed Futures itself could be classified as “niche” actually. The “success” of that ETF would probably detract from CTAs business, but rather in terms of AUM than performance, so I do not think this is a self-defeating idea from that angle…

    @Pretorian: re. 1 and 3: the index was developed by Thayer and Brook according to the press release. Never heard of them before but they are systematic trend followers in futures (according to this page: this page).
    re. 2: true and similar to the point made earlier with USO, I believe. Not sure if they would have to publish the exact rules of their system. It appears that Meb Faber’s GTAA ETF rules are not public for instance.
    re. 4: Thanks for this. Another product that I was not aware of myself.

  • Managed Futures: Alpha, Beta and ETFs | AMFM Blog

    [...] was a recent post at Au.Tra.Sy, that really hits to the heart of what we do at Auspice.  It relayed announcement that S&P had [...]

  • Auspice Tim

    As a CTA managing both alpha CTA products and “third generation” enhanced beta CTA indices, we do not see it as a threat. We believe in the separation of the alpha and index methods. We feel it will bring awareness to CTA and offer product to investors that do not know about/understand or have access to “Accredited or QEP” products. For CTAs that generate Alpha at reasonable price, this exposure can only help you. We are in this camp.

  • Auspice Tim

    Re Humble System Builder comments:
    The Claymore ETF you speak of never launched. Auspice is indeed the index developer and manager. The Broad Commodity Index linked CBR ETF did launch last October. A Managed Futures ETF linked to the AMFTRI index (launched and published NYSE in 2010, see our website) will likely launch later in 2011.

  • HumbleSystemBuilder

    Much apologies Tim, I was incorrect.
    I made an assumption. I read all about the launch of it and noticed the articles all had mid to late 2010 dates and noticed material on your website discussing it… then I couldnt find it on Claymore’s site or get a quote. I only found that the Broad Commodity ETF was actively trading. I should not have made the assumption that it was discontinued when it simply had yet to be launched.

    I follow the index currently and read your methodology on the index. I think it is great that you are taking measures to promote managed futures in Canada and abroad. The asset class has a long way to go in terms of recognition at the retail level especially in Canada.

    I very much respect your firm and have been following your performance and product development. I wish you all the best in your future ventures.

  • Jez Liberty

    Tim, Thanks for dropping by and chipping in with your precisions on the Claymore ETF, as well as your view on alpha vs beta products in the managed futures space. This is a good point actually.

  • William F.

    Do you know anything about the system they intend to implement in their trend following.

  • Reformed Trader

    Yes, for most retail traders, I think the potential underperformance of ETF’s compared to actual trading strategies and CTA’s won’t be an issue. Beta actually looks like alpha when you compare yourself to someone who has little diversification, which is almost everyone in the retail investing space.

    By just adding some exposure to commodities and REITs and adding a trend-following filter, a retail trader can handily beat the “market,” even if all he has added is beta.

    Of course, more is better, but there is going to be a point for each of us where we stop and say, “enough.”

  • Reformed Trader

    After reading through these comments, I decided to write a post to the average investor arguing that he should gain more exposure to non-traditional forms of beta.

  • Jez Liberty

    @William: no, I have only seen the press release and no further details. I am a bit intrigued by their “adpative” approach and wonder how much details they are actually going to disclose…
    If you check Thayer Brook website, they seem to be marketing a similar approach for their fund, so I would imagine they’d want to keep the logic fairly confidential.

    @ReformedTrader: yep, “exotic beta” products/funds (like this one) are probably a good way to beat the “traditional” market (which, as you say in your post depends of which benchmark you consider as “the market”). This blurs the line between alpha/beta a bit more, which is well illustrated by the A. Lo graph I used on the Betafication of Alpha: towards a Commoditization of Trend Following? post I mentioned in the post.

  • Reformed Trader

    Jez, your post that you mention along with something Bill Gross wrote got me started thinking that simple portfolio management methodologies can be used to great effect if we start treating trading strategies as assets.

    Have you seen this paper?

    Portfolio of Risk Premia

    I don’t like the paper’s long/short focus, as I think it is vulnerable to tail risk, but a modified strategy using swarm management would limit risk by trading a limited number of asset classes and limiting total exposure to each asset class to a manageable level. This is the direction I am going.

  • Praetorian

    Reformed Trader, I couldn’t agree more with you. I also think that traditional focus on equities and bonds or in this case equity long/short blinds a lot of people about different possibilities of building a portfolio of non correlated strategies, I have done some research with these approach and it gives a much better risk/return profile than using “assets” as diversifiers. By the way, if anyway is interested in this approach you might consider reading Veryan Allen’s blog, he introduced me to this concept.

  • Reformed Trader

    Thanks for the lead, Praetorian. I’ll check the blog out.

    The biggest challenge I’ve faced so far with trying to implement this strategy is controlling tail risk.

    As a simple example, consider trying to integrate merger arbitrage into a swarm management strategy. Since the strategy requires fresh exposure to stocks, you have to either ramp down your exposure to the existing assets or increase your total leverage.

    I make a distinction between total leverage and beta exposure because beta hedging may not be effective against tail events.

    Since merger arbitrage has zero beta exposure but increases total leverage, your portfolio becomes more and more centered around reducing tail risk exposure and reducing absolute return in order to do so.

    I’ve had the same problem trying to integrate value into my strategies.

    I’d be interested in hearing how you’ve decided to deal with these issues.

  • Reformed Trader

    To show a contrasting situation: suppose you’re trying to add momentum exposure to your existing portfolio. You can just take your existing assets and go longer the ones with positive momentum and go shorter the rest. So if you have a long position on SPY and your momentum strategy says to go short, the two strategies create perfectly hedged positions.

    Unfortunately, you can’t do merger arbitrage on SPY. So there’s the problem above.

  • Jez Liberty

    @Reformed Trader – thanks for the link to that paper; and thanks for the link to that blog, @Praetorian

    Using portfolio of strategies instead of/combined with portfolio of assets is definitely something I believe in, since you can “engineer” the strategies/systems to aim at lower correlations (which you can not do with instruments/assets). I actually mentioned this (Systems (And Timeframes) Diversification: Swarm Behaviour
    ) in a recent post:

    I actually started looking at LSPM (by Vince) with the idea of a a two-tier level allocation optimisation, first with assets, second with systems… But have encountered a few problems with the R implementation and have not investigated this further yet.

    This would probably not address your total leverage issue though.. Which sounds like a tough one and not something I have really looked. AllAboutAlpha had some good articles on tail risk management rcently. Maybe you could check their archives (archives are subscriber-only though…). I think Montier also had a recent piece on tail risk strategies (with a negative outlook on them I seem to recall).

  • Michael Harris

    Hi Jez,

    I suppose these ETF will be like traded funds. They will popularize trend following in some ways. I ‘m afraid they will cause shorter trend durations while increasing volatility and as a result CTAs will need higher winning rates. I have a quantitative explanation how this works here:

  • Jez Liberty

    Hi Michael,
    Yes, this “popularisation effect” might be a good thing for trend following in general, if it helps the concept getting more mainstream… I was actually reading an article very recently describing something similar in the coffee shops industry with the Starbucks effect – how new Starbucks opening up just next to independent coffee shops actually boost their sales:

    Soon after declining Starbucks’s buyout offer, Hyman received the expected news that the company was opening up next to one of his stores. But instead of panicking, he decided to call his friend Jim Stewart, founder of the Seattle’s Best Coffee chain, to find out what really happens when a Starbucks opens nearby. “You’re going to love it,” Stewart reported. “They’ll do all of your marketing for you, and your sales will soar.” The prediction came true: Each new Starbucks store created a local buzz, drawing new converts to the latte-drinking fold. When the lines at Starbucks grew beyond the point of reason, these converts started venturing out—and, Look! There was another coffeehouse right next-door! (full article:

    It will all depend on the buzz and marketing around the ETF I suppose.

    With regards to your note on increasing win ratio, I have to admit that I am not a big fan of separating Win/Loss ratio and Win rate to look at them independently for system design – as I feel they are two intertwined variables (change one will affect the other). Moreover, the increase in volatility during the 2007-2009 period you refer to in your post did not stop trend following to work. It was actually a fairly good period for trend followers, with the Barclay CTA Index and Newedge CAT Trend Sub-Index being up by over 20% in that period. As long as trend following can capitalise on outlier moves (ie fat-tail of the distribution), I believe it can afford a low(er) win rate because the Win/Loss ratio stays high.

  • Michael Harris

    Hi Jez,

    > With regards to your note on increasing win ratio, I have to admit that I am not a big fan of separating Win/Loss ratio and Win rate to look at them independently for system design – as I feel they are two intertwined variables (change one will affect the other).

    I can understand you but I have derived the relationship between win rate and R from the simplest possible principle (sum winners – sum losers >0) . Actually, they are not independent but one is the independent and the other is the dependent variable. Take any backtest report and you will see that if you know the ratio R and the profit factor you can calculate the win rate and that will agree with the reported win rate.

    Regardless, the real issue in my mind is the size that will be invested in those systems. When they exit a market they will create high volatility and as a result cause drawdown for CTAs.

    Not a good thing I believe. The more synthetic products you get out there the more room for market manipulation, higher volatility and even black swans.

  • Pumpernickel

    If you search not very hard you will find that S&P launched a Mechanical Rule Based Trend Following index (trading futures contracts) a couple of years ago. It was designed by “Market Wizard” Victor Sperandeo, and more than one ETF is available and trading, right now, using S&P’s publicly disclosed trading rules. To the googles!

  • RB

    I do not think this type of ETF will cause draw downs for CTAs or more market volatility. It will be for retail investors and they are long only and value investors. Financial planners will be reluctant to add it to allocations because commodities are thought to be speculative. It will be lucky to get half a billion under management. Contrast to David Harding with 20 billion. It will also have to put up the numbers to be able to get assets. Long term trend following has a lot of advantages built into it. Less trades so therefore less commissions and slippage. It has a fair amount of small losers and some small winners, and a few really big winners. Trend followers make some of their biggest wins in wild markets. Their is already trend followers using short term systems. I don’t like day trading or short term trading as this is the most random part of the market. I do see value in adding one to a long term system if it helps with drawdown. I think market manipulation has become a kind of cultural myth. Maybe in Chicago in the 1880′s . Markets go through cycles of volatility and I do not believe manipulation, trend following or HFT are the cause. The 1930′s had lots of volatility but no computers.

  • Jez Liberty

    Thanks for the pointers Pumpernickel. I have been doing some “google research” (not very hard indeed, especially with the clues you gave) and I believe you were referring to the same DTI index that Tim Pickering was also referring to above and in his blog post. This index does have a few products based on it, with about $3 Billion AUM over all products.

    To save other people from having to do their own research, I have summarised what I found in an update #2 to the post above.

  • Praetorian

    Reformed Trader,

    The way I try to manage tail risk is to use it in my favor, this means that the core of my portfolio is Trend Following which have positive skew and produce huge returns in case of tail risk events. I use other strategies to hedge or rather smooth my equity line, sadly these strategies are not as robust as Trend Following, it is very difficult to find CTAs with good track records and good MARs around 1 for long periods that are not TF. In any case they are still very good when blended with the core, I have seen on this blog papers that suggest that even strategies with negative mathematical expectation can improve portfolio MAR if the are negatively correlated.
    I have thought about using traditional Hedge Funds in my strategy like merger arb and long/short equity. However in my interviews with them I find myself not liking the way they control risk by not using stops and buying value and things like that, I guess I am too much of a Trend Follower and I have my limits, hope this helps.

    Could you explain the term Swarm Management in the trading context? or lead me to some info. I am reading a book called “The Perfect Swarm” but I find myself unable to find the applications yet (I am sure there is something).

  • Reformed Trader


    The combination of strategies I used is protected against tail risk when it’s accompanied by strong momentum. Most tail events are preceded by momentum: 2000-2, 2007-9, for example. I agree with you that trend-following is the best hedge in these cases, with global macro somewhat helpful as well.

    There are some instances of tail risk that were not accompanied by much momentum: for example, the crash of 1987. Some shorter-term trend-following strategies would have picked up on the negative momentum, but longer-term ones like Mebane Faber’s 10-month moving average would have been too late.

    I’ve hedged the latter kind of tail risk by keeping control over absolute leverage. Even if all the stock markets in the world drop by 30% tomorrow (unlikely, but possible), those who aren’t too badly leveraged with live to play another day.

    I was already implementing a version of swam management before running into Jez’s blog entry. The idea is to increase leverage while keeping absolute leverage under control.

    One example is to pick five ETF’s to trade that are normally not too correlated to each other. I decide on some maximum amount of leverage and then aggregate trading strategies around these assets. If all my strategies tell me to go in the same direction, total leverage will never exceed the maximum that I’m willing to tolerate.

    Since the strategies oppose each other most of the time, I’m pretty leveraged in theory but well hedged most of the time in practice. In the recent downturn, I was short Brazil and had reduced exposure other asset classes.

    The key to reducing absolute leverage is to make sure that all strategies trade only the a small basket of ETF’s. Unfortunately, this eliminates good strategies like value and carry trading, but it’s not a problem I’ve figured out how to solve yet.

    I hope this was helpful and not too simplistic of a strategy. I want simple stuff that won’t have a tendency to break, but I may be giving up some return in order to do so.

  • Jez Liberty

    Hey Prateorian, did you check the links to the Farm River “swarm behaviour” papers from my other post (futures with a asmall account: ). These were actually pointed out to me by Pumpernickel (on another blog) and do a pretty good job of introducing the concept in a trading context.
    part 1 and part 2

  • RB

    Jez ……Found another fund called the Arrow managed futures trend fund. Uses the Trader Vic index.

  • Jez Liberty

    @RB: Thanks, good “catch”.. I updated the post (again) to add a reference to this fund/index.

Leave a Comment