Research

Interview with Dr Ewan Kirk, CIO of Cantab Capital Partners LLP

Cantab Capital Partners (CCP) is a systematic global macro hedge fund that manages assets in excess of $800 million. The firm is based in Cambridge, UK and it has close ties to the University of Cambridge. Founded in 2006, CCP is a team of 21 mathematicians, computer scientists and finance professionals. The firm’s success can be attributed to the high-quality team, robust models and institutional quality systems and software. 

  1. Over the last two years, the performance of the CCP Quantitative Fund has been contrary to the rest of the market. The fund posted impressive figures in 2008, while in YTD 2009, it has been in the red. Could you shed some light on these notable trends?
  2. It is certainly true that the CCP Quantitative Fund has underperformed the equity markets in 2009. Since we are absolute return managers rather than equity managers, there is no reason to believe that we should outperform the equity markets every year. We are, by design and delivery, uncorrelated to the equity markets and that means that there will certainly be years where we are down when the equity markets are up. Of course, the negative returns this year should be put in context. Since our inception, we have returned more than 50%, whereas over the same period, the equity markets have lost 26%.

  3. Unlike many single-model managers, you run a large diversified basket of models. Which have performed the best and which have been the biggest drags on performance this year? What is the typical breakdown of your regional and strategic allocations?
  4. Our value-oriented models in currencies have performed well this year. These models pick up carry style risk premia which have performed very well. Our dynamic risk allocation systems have gradually upweighted this over the course of the year as a more “risk on” environment has emerged. Our momentum models lost money in March and April this year in line with many trend-following models. Given the sudden and dramatic reversal of many markets in Q2, it is not unexpected that these models will give back some of the gains they made in 2008.

    The allocation between each asset class and each model is rebalanced on a daily basis to ensure that we maintain a constant and optimal risk profile. Typically, this is an allocation of 20% to commodities, 30% to FX, 20% to rates and 30% to equity indices, but these allocations of risk can vary considerably as risk changes in each market.

  5. What is your view on why global quant and global macro strategies have not performed well in 2009 YTD?
  6. The main issue for most systematic (and discretionary) managers has not been the lack of trends but the fact that during the first half of the year, the general risk appetite environment was particularly unsettled. We and many others suffered from flipping between “risk on” and “risk off” relatively quickly. This was due to considerable uncertainty surrounding whether or not the nascent recovery was actually the beginning of a long and slow grind back to normality or just a bear market bounce.

  7. CCP is a fully systematic fund, which develops and deploys sophisticated mathematical algorithms to trade. Could you please tell us a little about these ‘sophisticated mathematical models’? How do these models predict and protect the fund against negative fat tail market events? 
  8. Obviously, many of the details of the algorithms are proprietary, but in general, each model is supported by an underlying economic rationale which drives the returns. This can be, for example, a momentum, mean reversion or value orientation at a particular timescale. These models generate the signals, but it is our portfolio construction and risk management models which help to minimise the potential for unexpectedly large losses. The portfolio algorithm uses a variety of innovative techniques to forecast the overall portfolio volatility and scales the portfolio appropriately. However, this is augmented by an entirely separate tail risk management process which reduces the size of the portfolio when the conditional VaR rises above a predetermined limit.

  9. You are constantly engaging in new research. Can you share how this helps in strategy development? What are the criteria of choosing and discarding strategies on an ongoing basis?
  10. Quite fundamentally, research and development is the main activity at CCP. Our infrastructure for prototyping, testing, evaluating and implementing strategies is second to none and this enables us to rapidly and robustly convert a theoretical hypothesis into a fully functioning live strategy. The criteria for choosing strategies are more subtle than “it works”. We have to look at the contribution of the strategy to our overall return stream and also how the trading of the new strategy will interact with the existing strategies. On an ongoing basis, the live strategies performance is compared on multiple dimensions to the expected performance. If a strategy does something which is “sufficiently unexpected” then it is removed. This is a systematic and robust process that enables us to remove strategies which have either been unlucky or for which the underlying hypothesis has failed or has degraded sufficiently.

  11. Can you explain the differences between the various classes of funds managed by CCP? How do they meet the different requirements of investors?
  12. We run a single fund internally, but because investors have different volatility targets or liquidity constraints, we wrap the basic returns in a variety of share classes. The basic share class is the Aristarchus which has a monthly liquidity and targets 20% volatility. This share class accounts for more than 80% of our assets. We also have a lower volatility share class Babbage and a daily liquidity class Dirac. The Aristarchus class is also available in a sterling hedged format. If investors have a particular requirement, we encourage them to discuss this with us since it is a minimal overhead and cost to create a new share class which wraps the basic Aristarchus returns in a different format.

  13. Often, it is difficult to describe complex quant models to investors. Have you found this as a hindrance to raising capital?
  14. Actually, most of our investors have been pleasantly surprised with the degree of openness that we show about describing our models; thus, we have not found this to be an issue on capital raising. As we often say, it is even more difficult to describe the internal workings of a discretionary trader’s brain and that does not make it difficult for them to raise money. In comparison, systematic models are generally simpler to understand.

  15. How has asset growth been for the fund over the past nine months? What is the rough breakdown of your investors by type/geography?
  16. We started the year with less than $500 million and are now managing in excess of $800 million. It was particularly pleasing to raise money during late-2008 and early-2009 when other funds were suffering redemptions. We believe that our very liquid underlying securities and monthly liquidity with no gate allowed investors to feel comfortable that they could redeem money if they wanted; and therefore, were happy not to redeem, or in many cases, to add to their allocation.

  17. How did the volatility impact the functioning of your funds? How do you foresee the amount of stress laid on the tightening of regulations across the hedge fund industry to affect your fund?
  18. Our fund targets a constant volatility and so, as the volatility last year rose, our position sizes got commensurately smaller and thus, the impact of the volatility was relatively muted. The regulatory environment is very fluid at the moment and seems to be dominated mostly by speculation and rumour and therefore, we are waiting to see what proposals emerge (if any) before considering how they might affect us.

  19. Do you expect your performance or investment style to change going forward? Could you tell us what are the market opportunities that you are looking at right now?
  20. Our process is fully systematic and therefore, we do not expect the style to change. Obviously, new models will be incorporated over time, but we fully expect to continue to generate high returns which are uncorrelated to the equity markets. Our investors are those who value the lack of correlation that CCP has in comparison to the other assets in their portfolio and we believe that our focus on this should be valuable going forward.

Contact Details
Dr Ewan Kirk
Cantab Capital Partners
+44 1223 755755
info@cantabcapital.com
www.cantabcapital.com