85% Hit Rate Pays Off For CTA Warren Capital

By Kamran Ghalitschi, publisher of HedgeNordic

In another year turning out to be challenging for Managed Futures, the Nordic Hedge Index CTA sub index just managed to creep into positive territory for the year, largely picking up some tailwind from trends in equity markets. Warren Capitals two CTAs outpaced the index, clocking in the better side of 8% return for Fourth Moment Macro and over 11% for Warren Short Term Trading, with a remarkable Sharpe Ratio of 4,2. Both funds were awarded at the 2012 Nordic Hedge Awards, Warren Short Term Trading as best Nordic CTA and even taking second prize overall as “Best Nordic Hedge Fund”, behind Asgard Fixed Income fund from Denmark. We took the opportunity to talk to Peter Warren, the companies CIO and founder about his trading approach, and here is what he had to say:

Kamran Ghalitschi: Congratulations on yet another great year. How satisfied are you with results compared to previous periods?

Peter Warren

Peter Warren

Peter Warren: Actually I am a bit disappointed. One of the reasons was predictable as we chose to spend a lot more time on generating and testing new ideas than the previous two years. As we have not had direct interface between our models and our execution system, less time trading – all things being equal – reduces revenue. Parts of our efforts have however gone into building a direct link between our signal generation and our execution platform. We plan to go live with this in February and expect through this to be able to keep a lookout for opportunities at all times.

The second and more justifiable reason for being disappointed was that we experienced a larger average drawdown per trade compared to 2012. It took time and quite forensic investigation to uncover the reason for this. It was so embarrassing simple that I will refrain from telling you what it was. Needless to say it has been corrected.

Kamran Ghalitschi: As rule of thumb, 2013 CTA results could be described as “the more exposure to equity markets, the more biased towards the larger markets such as S&P 500 and the more biased to slower models, the better the chance of landing on top of the tables.” How true is that for Warren Short Term Trading (WST)? Where did you pick up performance and where did you loose some?

Peter Warren: Actually our returns from trading equity indices ranked third in terms of performance. Fixed income came first followed by currencies. Regarding the S&P-500 our observed correlation versus the returns of this index has been 0.07. Looking across the asset classes we trade in, the highest observed correlation is 0.27. Regarding the slow versus fast models few would regard intraday trading only as belonging in the “slow” category.

Kamran Ghalitschi: Another observation from 2013 trend models seemed to have been working better than non-trend. What are WST key drivers for pattern recognition?

Peter Warren: Given your statement it is interesting to observe that our non-trend models were more active and profitable than our trend models in 2013. I guess this suggests that we, for what it’s worth, are doing something different than most of our peers.

Kamran Ghalitschi: Can you tell us a little more about WSTs basic trading approach and philosophy?

Peter Warren: We work under the premises that there is little or no informational value to be gained in the intraday timeframe where it is we operate. There obviously is information floating around in the market place but we do not have access to it, nor does anyone we know on a systematic basis. Based on this we believe that all that is on the screen in front of us is “noise”. Prices are moved by random and temporary movements in supply and demand. They are also influenced by rumors as well as scheduled and unscheduled events. We treat all of this as “noise” and apply non biased mathematics to generate the signals.

Given that we do not “know” what can affect prices at any given time, risk control is essential. We do not care or have time to find out why something is moving against us. If it hurts beyond X we just get out – quickly. You can have a coffee or beer later and talk about the why, who and what. In the moment all you should concentrate on is cutting the loss.

Kamran Ghalitschi: Risk adjusted returns as well as Sharpe Ratio are unusually high for WST, while volatility is remarkably low for a CTA. Especially considering the period since inception of WST were three down years in Barclay Hedge CTA index. Is that by design, or have markets played into your hands?

Peter Warren: To be honest we are a little bit overwhelmed by the risk adjusted returns being this good. We have always been strong believers in evaluating returns according to risk and would love to tell you that this level of Sharpe is what you can expect us to deliver for the next decade. However this is too high a bar. That said, having on a number of uncorrelated positions with a mathematical edge does a lot to keep your volatility down.

Kamran Ghalitschi: In your monthly letter to investors for December, you write: “research has uncovered more markets that meet our standards in terms of liquidity and we expect to have some of these included in our next release of trading algorithms.” Can you let us in a little more what markets you trade now, which ones you may add and the advantage you see in further diversification.

Peter Warren: Up until now we have only traded in twelve of the most liquid currency pairs and the most liquid futures contracts (Mini-S&P, Eurostoxx, DAX, T-Bonds, T-Notes, German bunds, gold and oil). It is ironic that we trade where competition is at its fiercest.

We are currently looking to add more currency pairs but really we can add anything providing it has sufficient liquidity and low trading costs.

Kamran Ghalitschi: How much effort are you putting in further research & development? How much of “if it ain´t broke, dont fix it” is true for a trading model?

Peter Warren: We think about markets and their behavior a lot and constantly test new ideas. However we try to leave models that work alone and concentrate our efforts on the next “great thing”. It is worth noting that our models are adaptive and not static. They pick up a number of factors during the day and adjust accordingly. We believe among other things that changes in volatility and/or kurtosis are important factors to watch.

Kamran Ghalitschi: The same release also warns investors, not to be expecting risk/return numbers to remain and those levels. Where would you reasonably expect Sharpe Ratio for WST to settle down in the long run?

Peter Warren: 1.5 to 2 would be a great range over time.

Kamran Ghalitschi: You also pointed out a good hit ratio for WST. Can you let us in what your ratio of winning to loosing trades is, and how much you gain on a winning trade and loose on a loosing trade on average?

Peter Warren: Our hit ratio has been fairly constant at above 85. We would expect this to come lower when our trend models become more active. The non-trend models typically on average have larger losses than gains. The ratio, as mentioned earlier, had widened to 2:5 (still providing us a positive expected return) before we realized there was something amiss.

Kamran Ghalitschi: There has been much discussion and research published how smaller, emerging CTA outperform the larger peers. How do you see the relation of AuM and being in a position to produce excess returns.

Peter Warren: Becoming a behemoth like Lynx and the Wintons of the world excludes you from being nimble and therefor from a number of trades that we find attractive. This is in a way good as I believe we create opportunities or do service to one another.

One of the reasons for building new models and expanding our instrument coverage is to be able to increase capacity as well. It is hard to know what our limit in terms of AUM is and have therefore designed a mechanism for controlling the size of the fund by annually paying out all profits accumulated over the previous year. We have had a soft lock on the fund for the past 6 months and only now beginning to let new money in, albeit at a controlled pace. Currently we are fortunate enough to have a growing list of investors who would like to come into the fund.

Kamran Ghalitschi: Best guess is, Norway does not offer you the perfect environment for running an alternative investment approach. How much of a handicap is it being domiciled in Norway, and how do you see the future?

Peter Warren: We are handicapped in the sense that not many foreign investors make it to Norway to do due diligence. We are also handicapped by Norwegian regulations being heavily in favor of protecting the mutual fund business from competition. We have not even bothered to apply for a license to sell WST to Norwegians as our other fund Fourth Moment never was given the advantages promised after we went through the process and got a license for this fund. Other than ourselves there are only non-Norwegian investors in the fund. WST remains domiciled in Ireland and regulated by the Irish authorities.

Kamran Ghalitschi: Before setting up Warren Capital, you had a military career in the British Special Forces. Is there anything you can take from that time and put it to go use at Warren Capital?

Peter Warren: There are two things that I believe stick with you after such an experience: discipline and to never assume.

Kamran Ghalitschi: Can you share a piece of market wisdom with us?

Peter Warren: Find a way to trade that fits your personality.


Reproduced with kind permission from HedgeNordic (http://hedgenordic.com)