System Trading Blog

Make Your Money Work Harder

Currensee: premium fee for basic service

leave a comment »

Image Currensee allows investors to mirror trade real trading accounts of traders. As a part of service, they do due diligence on traders before they approve them as eligible for investor accounts (and brand such traders as trade leaders). This is a big difference from other services like Collective2 and Zulutrade which are open to all traders. But then, the choise of which traders to select is up to the investor himself. Regulatory and corporate liability considerations should play a role here. They charge a hedge fund like fee of 2/20 fee (2% management fee and 20% incentive fee). Given the fact that investor selects the traders himself, this is a high fee for the service (additionally, Currensee receives rebates from broker).

Currensee pays the traders 15% of the profits they make for investors, but no share of management fee or broker rebates. This is a near-optimal arrangement as the investor interests are aligned with those of traders and there is no incentive for churning. And this is a one of the best deals out there for traders.

Currensee, now owned by retail forex company Oanda, is one of the safest ways to start copy trading, but it is heavy in terms of fees and for anybody with basic skills in trading algorithm evaluation it is more beneficial to look elsewhere. The reported average return of 11.94% p.a. (as on Sep 30, 2013), not adjusted for fees, is less than impressive for this type of investing.


Written by A.S.

October 29, 2013 at 10:29 pm

Movie about old school trading

leave a comment »

trade.the.documentaryDigged up an old movie Trader – The Documentary about early days of Paul Tudor Jones. The first 40 minutes is a very illustrative example of how trading was done in 80’s (and continued till late 90’s). The movie is a kind of classic trading movie.

In those times the slippage and transaction costs were much higher than today even if adjusted for the “evils” of high frequency trading.

No copyright information. Probably was done as PR movie. Just google this for download: []trader.the.documentary.paul.tudor.jones

Written by A.S.

October 28, 2013 at 6:34 pm

Is high frequency trading creation of evil genius? [updated April 20, 2014]

leave a comment »

HFT“There has been in our culture, in the past decade in particular, a group of reasonably smart people who hired incredibly smart people – mathematicians mostly – to design algorithms that exploit time/space phenomena such as latency to vacuum insane amounts of money out of the economy, for doing absolutely nothing except exploit systemic flaws in the digitised financial world. We’re talking about hundreds of billions of dollars, if not trillions, simply for hiring bright grad students, hurling some cash and some lap dances at them, then hitting the return key and making a billion dollars in a wink of an eye.”

Full article, as published in FT, is here.

Some more accounts on nature of HFT are here and in more detailed way here (particularly, the part from 7:55 to 9:00 about qoute stuffing and the rationale behind qoute stuffing: it is easier to make noise than to make meaning from the noise).

No doubt, the reports in both literary and technical versions have a substantial degree of accuracy. There are investment pools out there. Often closed to outside investors and often generating astronomic returns and, yes, involving bright quants; and different groups often dominating different exchanges/markets/countries.

Is this all evil? Of course, not. First of all, there are markets where this does not work, forex as an example. Then, no need to revisit the shortbacks of open outcry — the total transaction costs, and slippage cost as a separate item, were much higher. The exchange systems improve and, as time passes, HF groups will bring each other beyond the point of profitability.

What will remain is efficient, safe technology allowing for low transaction costs. Anyway, none is going to cut 2/3rds of trading volume, i.e. the share of volume HFT accounts for in many markets.


Update on April 20, 2014:

I´ve been blissfully uninformed when I wrote the initial post. Looks like there is a bit of evil in HFT. But just a bit. If it is true that the whole stock exchange system was redesigned by HFT operators to gain priviliged access to market data enabling them to skim majority of market participants (i.e. 99.9+% of them) in a systematic way, I would not say it is all good. However, transaction costs are tens of times lower than they were before trading moved from open outcry pits to servers. In that context, the investors are still better of and the fact that HTFs are skimming us is a minor side effect.

Nevertherless, even this side effect may be a thing of past as it has attracted some attention and is likely to cease. Let’s see if they dig up something in this process. And there is a good literary account how HFT process works:

[watching the same symbol in private account and on Bloomberg terminal]  ¨In his private brokerage account he set out to buy an exchange-traded fund … ¨I hadn´t even hit Execute¨ …  ¨I hadn´t done anything but put in a ticker symbol and a quantity to buy. And the market popped.¨ (Page 67, Flash Boys by Michael Lewis)


Written by A.S.

October 27, 2013 at 11:07 am

Managed futures getting obsolete

leave a comment »

Returns of managed futures are at generational low. This is not a cyclical event; it is a structural change. Managed futures are getting obsolete. Technology has reduced transaction costs and data feed costs manifold. Futures, alongside with stocks, forex, ETFs have become accessible to virtually any investor directly. One can trade and execute strategies on electronic platforms bypassing any human contact, like broker or advisor. They can tail-ride traders, investors or signal providers on services like, or More and more investors do this changing markets in the course.

Consequently, the price behavior has become more fragmented. There are no 1970s type trends that can be captured by simple, often including elements of pyramiding, trading strategies (like one I have written about here). Today successful strategies are more specific and with shorter time span. Traded by an insider of any specific industry (biotech, industrial machinery etc.) or some technies in high frequency trading aiming to access signal a few milliseconds faster etc. Due to fragmentation, a single manager (and this includes institutional managers) is too limited in scale and scope to benefit from such more competitive niche strategies. For typical money manager many top-performing niche strategies have too limited market size (say, how US$25 billion Winton Capital Management can profitably capture a top-performing strategy with US$ 50 million capital deployment limit) or too short life span (can a typical hedge fund make a complete strategy overhaul every 3-6 months) or lack of institutional capacity to manage large number of diverse strategies (there is a limit on how many strategies one person can develop and maintain); and this is even before talking about legal restrictions of offering memorandum.

There is little managed futures can offer in terms of cost improvement (everyone can trade cheap on virtually any scale) or access to investment (even a small investor can trade all the spectrum of futures), so managed futures are even in more competitive area than mutual funds or hedge funds. But it is clear that the key drawback comes from the very asset management structures with their inflexibility (due to both legal and organizational issues) in regard to strategies employed, an inflexibility that renders managed futures (and all other mainstream investment vehicles) unsuitable for today’s markets.

Disclaimer: We’re working on development of a web-based service to enable sourcing and dynamic application of trading systems to meet the challenges of the modern marketplace. The traditional business entities like banks, mutual funds and hedge funds are not fit anymore to provide such investment service with consistent profitability to investors.

Written by A.S.

October 2, 2013 at 11:28 am

Posted in managed futures

Tagged with

Bitcoins: a version of tulipomania

with 4 comments

ImageBitcoin value will crash, even if it goes through bubble phase of tulipomania scenario first. Bitcoin (or BTC) is worth something just because of market participants believe it is worth something. Why people keep buying BTC? Fun of participating in something novel plays a role for some. Greed should be present as the price increase looks like no-brainer to others. Certainly some of BTC value derives not from what it is, but rather what it is not.  What it is NOT: is not a fiat currency that may devalued by “printing press”, authorities can’t [yet] track down or block every single transaction, storage of the value is not dependent on solvency of any particular institution etc.

Video: You Really Should be Buying Bitcoin (a good explanatory video on BTC)

Without any central institution to destroy the value, no oppressive body to track and/or tax it and, generally, ease of transactions, for many BTC certainly is an attractive alternative to money. At this stage, BTC can perform the essential functions of money — medium of exchange, store of value, measurement of value, no question about it.

There is one trait what sets BTC apart from any other money equivalent: the amount of BTC in circulation capped at 21 million. Even reaching this limit will take some time, as the cost of mining BTC is already higher than their current market value.

We can certainly imagine that in near future an infrastructure is in place to make transactions easy. Easier than cash. Like to make payment with NFC phone and Mastercard Paypass installed, for example. In fact, the technology is in place already. It is just the matter of adaptation by users.

So with all the wonderful specs why BTC is just another bubble? Something free from malaises of fiat money, something quantitatively more limited than gold or diamonds, as anonymous [yet] as gold or diamonds, no storage cost, ease of transaction and at price what is less than production cost.

The reasons BTC as a currency are doomed:

  1. Non-debaseable money – true, but it is not unique and for BTC serves no purpose
    The number of BTC is limited by its algorithm at 21 million. However, so can be any other crypto currency (like litecoin), it is a question of mathematic algorithm. To a certain extend non-debasable are gold, silver, land etc. as well – all of which have gone through boom and bust. What is worse, 21 million BTC in circulation is a setback for development as money. If BTC was to replace all the fiat money in circulation, any single BTC would be worth millions (imagine global GDP of about US$71 trillion to be serviced by 21 million of BTC in circulation) and price of cup of coffee would be expressed as 0.000001 or less BTC – not very practical.
  2. Easy replicable by other crypto currency algorithms
    It is a crypto currency. BTC is based on one specific mathematical algorithm. Similar algorithms are easy to create. Litecoin is one example. Other crypto currencies may appear in endless variations. Such other crypto currencies are likely to be suitable for transaction using the same infrastructure as BTC. For all practical purposes, if somebody creates a crypto currency with global ambition to replace fiat money, for all practical purposes it is better to create something with 71 trillion rather and 21 million units in circulation.
  3. Anonymity will disappear
    Several governments have moved to bring BTC under regulation. US has recognized it as “virtual currency”, Germany as “financial instrument” – whatever the name, in any case BTC is an asset, which is subject government surveillance for anti-money laundering, licensing and tax purposes. All BTC transactions leave a trail on internet. BTC anonymity is not there to stay. In a sense, BTC is a victim of its own popularity as it attracts government scrutiny.
  4. Momentum will fade
    The momentum of novelty will fade. Probably most people on retail level get involved in BTC for fun. This factor will wear off eventually. Speculative interest will disappear after first crush. If all that backs BTC is confidence, then confidence may easily be highly volatile factor. Particularly when people recognize that behind BTC there is nothing else but the confidence.
  5. Technological risks
    The technology of mining, transacting and storing of BTC is new and critical flaws may appear.

All said, the key lesson from BTC is that interest (demand, need) for an alternative currency is strong and I expect that some properly backed, non-national currency will be invented.

Is it a sensible thing to invest money in BTC? I do not think so. The thing of BTC funds with 1000% annual returns may be the thing of the past. Have people made money of it. Absolutely. Early miners, for example. Today mining is not profitable. Will value of BTC increase? No idea, but tulipomania scenario is quite probable. I guess, there are people who know how to trade through bubbles. Is there any practical value of BTC? Apart from entertainment and promoting a novel idea, BTC actually can help to move funds cross-border below government radar scrutiny (something one cannot do with any major national currency). At least by now, but knowing government dislike for it as showed by Liberty Reserve case and having US$11b budget and 35k persons to go after it, it is unlikely the loophole will persist for long.

Written by A.S.

August 30, 2013 at 7:47 pm

Is trend following still alive?

leave a comment »

Attain Capital Management has put out a quality piece of research named Is Trend Following DeadBasically, the argument goes that trend following programms are in noticable drawdown and historically this has been a good entry point to invest in such programms. And an implication that nothing has changed, trend following is live and kicking and as sound investment approach as ever.

I would like to add shades of grey to plain black and white picture. Obviously, trend following is not dead. Many traders and asset management companies still use it as one of their principal strategies. Then there is the issue of the definition of trend following which is vague at best. Ed Seykota, probably the best known and the most respectable trend following apologist, says that the difference in price necessary to create the profit implies a trend. But, hej, this is true about any investment approach, not just trend following.

In Attain’s case the question of the definition is more straighforward — they consider the investment programm as trend following if the asset manager labels it as such. However, the question remains open what exactly are the  trend following strategies those asset managers use. Technically speaking, the only difference between high frequency trading and trend following may well be just timeframe. None provides a detailed description of the models used and  those may well have evolved beyond recognition from the original ones. In that sense trend following may not be dead but the inner workings of the system may well be a completely different from the classic notion of trend following.

Then there seems to be sufficient evidence that the returns from trend following systems are declining. This is a trend in its own right. If to draw a trendline on the chart (see below) from Attain’s research, the returns are in obvious downtrend and this may well be a secular trend which is not going to change.


The Turtles are widely considered as the posterboys of trend following. They have been successful beyond doubt, but the returns from their system have diminished with time to reach a point where the returns are not competitive anymore. I’ve written about it before here.

Trend following is not dead. Traders still widely use it. But the inner workings of the system have changed over time beyond recognition and trend following today may well be something totally different from what it was decades ago. In terms of performance, the “classic” trend following has stopped working long time ago as can be seen from “the Turtle system” test results. And the chart from the Attain’s research may well serve as another evidence that the returns from trend following systems are on secular “downtrend”.

The trend following is not dead but in its classic form has decayed into useless and the modern version of trend following may be  different from other investment approaches in name only. However, nothing said here invalidates Attain’s suggestion that a bet on trend followers may be a good move now. Those asset managers are smart and rebound is highly probable.

Written by A.S.

November 7, 2012 at 8:30 am

How long it takes to launch a new investment fund in EU

leave a comment »

It should be around 5-6 months, providing that the founders and managers are acceptable to the regulators. It takes about 1 month to prepare documents for the regulator and they’re likely to use most of the 6 month period  allowed to them for decision to approve or deny licence.

It took us about 18 months. But we started at the very grass roots level and there were a lot of issues to consider, not least how to ensure client confidenciality. Most of our clients are based in Europe and the legislation is likely to change so that it is very difficult to market offshore funds in Europe. So the choice was to register the fund in Europe. The next decision to be made was in which country. Traditional fund domiciles like Ireland and Luxembourg were considered initially alongside with Switzerland. All of those are the highest cost countries for fund registration. Switzerland has an disadvantage of being outside European Union.

Our final choice was Malta for a number of reasons. It is EU country, the regulator is efficient (something we felt is true after the first meeting with the regulator there), business language is English and the registration costs are a fraction of what they would be in Ireland or Luxembourg. Malta currently is still a small, but the fastest growing fund domicile in EU.

There were thousands of email exchanges with prospective service providers (custodians, administrator, auditor, brokers …) and legal counsel. In typical situation this would be more simple, but in this case the fund principals did want to double-check every single detail themselves and a lot of time and effort was spent on understanding how even the slightest detail in the whole setup works.

I’m happy how it worked out this time — no mistakes were made, no cost overruns and no major delays (some minor delays were mostly due to the fact that big institutions sometimes are slow with new things). I believe the next time for the same people it would take 4-6 months to launch a new fund.

Written by A.S.

October 2, 2012 at 12:01 am