The Gateway to Algorithmic and Automated Trading

Accentuate the positive

Published in Automated Trader Magazine Issue 23 Q4 2011

Philippe Bonnefoy is the co-founder of Eleuthera Capital AG, where the investment philosophy is to make profits on capital when markets are not trending, and the belief is that this aim can be most effectively achieved by the use of automation. Larry Levy went to find out more.

Philippe Bonnefoy

Larry Levy: Philippe, welcome. Please tell us briefly about your background in the markets and how you came to automated trading.

Philippe Bonnefoy:

I started in New York in 1983, working on Wall Street in international corporate finance for Kidder Peabody. In 1986, I was asked to head up the international equity investment side of a new fund-management group they were starting. Two years later, I was invited by ex-colleagues to join a new risk-arbitrage group that evolved into multi-strategy, and then global macro proprietary trading group. At the end of the nineties, I began a relationship with Commerzbank, helping them grow their alternative assets. Then, in 2006, I sold my own alternative investment funds of funds to Erste Bank to concentrate on macro investing. I believed a new approach was needed, combining top-down global macro fundamental research with automation of short-term trading. For several years, the research we did yielded much better, more consistent returns in our short-term automated trading, so we decided to concentrate on that, particularly in FX. Currency investing allows you to express a number of ideas in a very efficient way.

Larry Levy: Your latest venture is Eleuthera Capital AG, of which you are the founder. Tell me about that.

Philippe Bonnefoy: I'm the co-founder with Riadh Fessi; I've worked with him for five years, he joined me from Credit Suisse. Eleuthera Capital was formed this year to operate our fully automated short-term currency-trading activities. It's an investment advisory firm. We have a new fund that's about to be launched, Eleuthera Currency Fund; we have managed accounts, and I also chair a London-based firm, Newscape Capital Group, which will be launching a UCITS version of our automated currency strategy.

Larry Levy: Could you expand on what attracted you to automated trading?

Philippe Bonnefoy: Automated trading is really the only efficient way to manage short-term time frames in the markets. I don't mean high frequency; typically our investment periods are from half an hour to twenty-four hours, sometimes a little longer. In markets that are moving, if you're trading more than one, you need to be able to execute trades efficiently, which means automating the process. The issue is, if you're trading multiple currencies, or multiple asset classes, in short-term time frames, by the time you can execute, you've already lost a bit of the opportunity you saw. The same applies to risk management: if the trade is not going to go your way, you need to execute efficiently.

Larry Levy: Do you keep trades over the weekend?

Philippe Bonnefoy: There's no preconceived notion that we shouldn't, but we don't often take trades home. It would be unusual for us to have many trades held overnight. The price action often calms down towards the end of the day, and that might be a condition for us to exit.

Larry Levy: What is your essential approach to risk?

Philippe Bonnefoy: We operate an investment methodology where risk is one of the core principles. Every time we instigate a trade, we already have a predetermined stop and a predetermined profit target.

As the market moves, hopefully our way, the stop will move into the trade at the point of entry and scale it up. We manage the risk so that the minimal amount of capital loss is expressed. The way we'd lose money is if, at the time of entry, the trade goes against us quite quickly, so we get stopped out quite quickly.

Larry Levy: So the stop moves consistently towards the trade entry.

Philippe Bonnefoy: Correct. We also have risk filters that moderate the amount of capital exposed to the number of positions. Each incremental trade would be sized smaller; that is also a constraint on risk.

Larry Levy: Is your approach, broadly speaking, based on technical analysis?

Philippe Bonnefoy: We only look at price action. We look at prices with a number of filters. The evolution over a number of years of developing the models led to an increase the number of filters.

Larry Levy: Do you have different strategies, or just the one?

Philippe Bonnefoy: We have two main models. One is short-term direction and the other is mean reversion. The essence of our investment philosophy is that we want to make positive returns on capital when markets are not trending. We're looking at intra-day time frames mainly, and we want to be able to take advantage of trade dispersion, of price action within trading ranges intra-day. If markets are active, trading ranges are wider, that gives us more potential for alpha capture; if markets are quieter, price action is more constrained, that reduces our capacity to make larger gains, but the whole essence of our trading philosophy is to make positive returns every day.

Larry Levy: Philippe, you've gone down the fund route rather than the managed-account route. Can your clients do managed accounts?

Philippe Bonnefoy: We have a managed account ready. Our preference is to run two distinct funds, one offshore, one onshore, but we do accept managed accounts subject to a USD 5-10 million minimum.

Larry Levy: So you have the Swiss Entity, Eleuthera Capital AG; the Cayman entity, Eleuthera Investment Management Limited, Eleuthera Currency Fund; and Newscape Capital Group, which is coming on for UCITS investments. What are the minimums across all those?

Philippe Bonnefoy: For the offshore fund the minimum will be USD 1 million; for the UCITS fund it will be very low - probably around USD 10,000. We hope to have the UCITS fund launched by the end of the year.

Philippe Bonnefoy

Larry Levy: You've launched a new fund with so far simulated results. What are your unique selling points?

Philippe Bonnefoy: We have a fourteen-month track record for incubation testing. That has been set up in order for us to make sure the models are working to our satisfaction, to test the technology, slippage, typically all the things you'd want to do before you took outside money. That has been highly successful and is up considerably and consistently; it fully validated our back-test results.

Larry Levy: Was that with real money?

Philippe Bonnefoy: Yes.

Larry Levy: Do you essentially use the same models that you used several years ago, or are you constantly evaluating and putting new ones into place?

Philippe Bonnefoy: The concepts of the models have remained consistent, but the filtering and constraining of risk, and allowing for repetition to make a return that is consistent; those are the key objectives we've had in refining the models.

Larry Levy: Your worst month - tell us about it.

Philippe Bonnefoy: It's still an incumbation test, but it was a minor loss of 50 to 150bp. That is typical of our profile, which is that we will expect to have drawdowns generated through a number of small trades being stopped out, which may then accumulate to a larger number, rather than one big drawdown from a trade or a theme. The way we would lose money is from sequential small losses.

Larry Levy: Where would you predict your possibly maximum drawdown would be?

Philippe Bonnefoy: We have planned, in how we manage the fund, to have sequential drawdown which could be, from peak to trough, somewhere between five to ten per cent. That would be within our expectation. We are capable, with our risk constraints, of making that number smaller, but it would also constrain the upside potential of the fund.

Larry Levy: Were we to experience another September 2008, would you pull the plug on trading, given the intra-day volatility?

Philippe Bonnefoy: The key drivers of our investment models are ones that like volatility and take advantage of wild price action. In our back-tested results, the period September to October 2008 was among our best ever. In real results in testing, similarly, periods of high volatility in the foreign-exchange markets, the months of November 2010, May 2011 in particular, with real money, were among our best months as well. The models do extremely well in periods of high volatility when other asset classes are suffering major problems.

Larry Levy: Eleuthera trades only currencies. What currencies, and what's the gearing?

Philippe Bonnefoy: We trade only the major developed-market currency pairs. We've taken the top eight currency pairs by volume, according to the Bank for International Settlements' data. That pretty much ensures that we have good liquidity. We have no defined level of gearing, in the sense that, when the market's more active, we may add more positions, which might add leverage. If the market is quiet, we won't. We might have periods when we're sitting on cash. We're not running a consistently leveraged fund; we are looking only to take optimal positions where we have high conviction.

Larry Levy: Do you trade during the so-called end-of-day period, which is 11pm European time, 5pm New York?

Philippe Bonnefoy: We have no constraints on when we trade. It's a quiet period in the market, so it tends to be quiet for us. But we could trade.

Larry Levy: So if there's suddenly an event at that time, your machine will cut into action.

Philippe Bonnefoy: Absolutely, yes.

Larry Levy: Why do you trade just currencies?

Philippe Bonnefoy: Because currencies have the best liquidity. They're also the best way for people to express macro views.

Larry Levy: Do you trade some currencies more than others?

Philippe Bonnefoy: We have no preconceptions on that. As the models are only analysing price action, they will typically trade the more actively traded currencies, which have more volatility.

Larry Levy: What technology do you use in programming and execution, and why?

Philippe Bonnefoy: We have spent several years doing programing work, and it's been a large part of the millions of dollars we've spent on R&D. Ultimately, we've found that MT4 was the easiest and best place for our style of trading and programming.

Larry Levy: Most people go for something more sophisticated, on a fund or institutional level, but some funds seem to come back to, and stick with MT4.

Philippe Bonnefoy: That's exactly what happened to us. We started off using very easy language for programming; MT4 was quite simple to utilise. During the course of our progress and development, we used some of the most sophisticated algorithmic trading consultants; we also had an in-house C++ programmer. We have used probably five to ten outside firms to help us in programming, and ultimately we found that the exercise was less efficient, particularly as we evolved the models to incorporate on-going research, than it was just to come back and make it simple for us to do with MT4. We've gone the full gamut, we've come back almost to where we started, and we're very happy with that.

Larry Levy: Keep it simple.

Philippe Bonnefoy: We felt it was right to pursue a more institutional infrastructure and development. In embarking on that pursuit, we found that it just made the whole process more clumsy. When you have more people involved, they make more mistakes, they have their own opinions and processes, and we were taken further away from where we wanted to be. So yes - keep it simple.

Larry Levy: Could you talk about your testing procedures?

Philippe Bonnefoy: It starts with an idea. We've been trading for almost three decades, so we have an idea of what works and what doesn't work in markets. As we look to add complementary features - that will reduce the correlation of the models to each other, say, or take advantage of different market conditions - we will first sketch the idea out, begin some model work, then do a back test for a year or two, then we may refine the model based on some initial research, then, if we're generally happy with that, which isn't often, we will run a much longer test. Once that's all happened, if we're still getting results over a much longer period of time, then we will begin putting some real money to work in a small test-trading account. If that's successful, we will incorporate it into our fund projects.

Larry Levy: Are you using MT4 for that back-testing?

Philippe Bonnefoy: MT4 and other systems.

Larry Levy: How do you quantify your risk/reward in terms of relative return?

Philippe Bonnefoy: We can dial risk up and down. The elemental way to do that is to decide what is the individual trade size that we want to do. Clearly, the larger the trade size, the larger the potential stop-out if we're wrong. Dialing up or down the position size of trades is the core way we can upsize or downsize the risk/return profile of the funds. In our back-tested research and in our real-money trading experience, we are typically getting an 80 per cent win rate on our trades. One surprising thing is, about our methodology and about currencies, is that the currencies themselves, intra-day, do not have particularly high correlation to each other. You expect they would, because an economic event, or a major speech, would have an impact across currencies. But what tends to happen is that the currencies react at different times.

Larry Levy: And this brings you back to the advantages of the automated model.

Philippe Bonnefoy: Absolutely.

Philippe Bonnefoy

Larry Levy: In this age of heightened risk, where does Eleuthera keep its client funds, and what is your fund structure?

Philippe Bonnefoy: We use a Cayman master-feeder structure. We only trade with the top financial institutions. If we have a managed account, the owner might have a view of where they want the money to be. In our case, we like safety, so our initial prime broker is Rabobank, which is one of the few triple-A rated banks left in the world. We do work happily with other banks where we have long and excellent relationships.

Larry Levy: Given your experience, is there any type or category of model that you prefer, or are there any that you automatically exclude?

Philippe Bonnefoy: I think we're open to all ideas. Having done this for a very long time, we do have a preconceived notion of what works, so that could be our human feature, if you like.

Larry Levy: Could you expand on your style of trading and how you came to this logic.

Philippe Bonnefoy: Our style of trading is about combining various features together. We look at market structure, time slicing of markets to see what's happening in certain incremental periods, volatility, trading ranges, pattern recognition, momentum, what's overbought, oversold. The essence of our models is to combine a number of factors together, not only to see what conditions persists in the market, but when conditions change in the market.

Larry Levy: Do you have a maximum drawdown loss-limit policy?

Philippe Bonnefoy: No, we don't. We believe that if you have an automated system, it should be fully automated unless there's some seismic shift. 100 per cent of our trades are fully automated.

Larry Levy: Are executions identical across all client accounts?

Philippe Bonnefoy: Yes. Unless a managed-account client provides a different set of instructions for us. Ultimately, it's their money.

Larry Levy: Tell us about your fees.

Philippe Bonnefoy: Standard industry fees, two and twenty. 2 per cent management fee, 20 per cent performance fee with a high-water mark.

Larry Levy: What are Eleuthera's back-office and execution arrangements?

Philippe Bonnefoy: The joy of automated trading is that you need very little resource in mid- and back-office because if you're doing it properly, you should be doing fully automated straight-through processing. That means it's very hard to make a mistake, if there's no human intervention, and you should settle cleanly into your prime broker without any human touch. The job of our mid-office and back office becomes one of monitoring to be sure that there have been no trade breaks, and of ensuring that the results we're seeing on the execution front end match the back-office record. We outsource fund accounting and administration to Apex Fund Services.

Larry Levy: Can clients withdraw money at any time?

Philippe Bonnefoy: In managed accounts, it's their money and they're directing us. They can withdraw authority to trade at any time. In the case of the funds, it's a much more controlled environment. We have weekly NAVs; you can withdraw your money on a weekly basis.

“The joy of automated trading is that you need very little resource in mid- and back-office because if you’re doing it properly, you should be doing fully automated straight-through processing.”

"The joy of automated trading is that you need very little resource in mid- and back-office because if you're doing it properly, you should be doing fully automated straight-through processing."

Larry Levy: Do you only trade your own ideas, or do you take on ideas suggested by others?

Philippe Bonnefoy: We are always researching in the markets. We talk to many people and we are always looking for new ideas. We think that we have perhaps the best models but we're always looking to improve them. We don't think that learning stops the moment you've made a model.

Larry Levy: What's your view on the future evolution of automated trading?

Philippe Bonnefoy: What's happening is a slight technology arms race. On the ultra-high frequency side, you have to spend an increasing amount of money to stay relevant and to stay at the front end of communication and processing speed. In terms of the more generic marketplace, which would include us and most market participants, all it's done is make the markets much more efficient in how they price, and it's lowered transaction costs.

Larry Levy: Where would you like to see Eleuthera in one year?

Philippe Bonnefoy: The asset-management market is often broken down into two modes of thinking. One is an investor-performance, net-return orientation, and the other is about asset-gathering. I'm in the former. I would like to see us delivering above our targets of 20-25 per cent return, consistently. Also, in the institutional marketplace, for better or worse, institutions want to see you manage a lot of money, because they want to feel that they're dealing with a robust counterparty. I would like our counterparties to feel comfortable with us, as they've felt comfortable with the companies I've had in the past.

Larry Levy: Finally, Philippe. We know you're quite sporty - what makes you buzz outside work?

Philippe Bonnefoy: I'm interested in sports and the arts; my father was an art dealer. I like museums and galleries. I'm a skier and a sailor, so when I have time, which isn't very often between school events and family life, I like to do those things. For the past few years, I have to tell you, I've spent most of my weekends working.

Larry Levy: Philippe Bonnefoy, thank you very much.

Philippe Bonnefoy