Seven Questions To Ask Your Systematic Fund Manager. Contents INTRODUCTION... 3

Transcription

1

2 Contents INTRODUCTION ) WHAT MARKET AND ACADEMIC EXPERIENCE DO YOUR RESEARCHERS AND PORTFOLIO MANAGERS HAVE? ) EXPLAIN YOUR STRATEGY IN TERMS MY GRANDMOTHER WOULD UNDERSTAND ) ON AVERAGE HOW MANY INSTRUMENTS DO YOU HAVE IN YOUR PORTFOLIO? ) WHAT DO YOU EXPECT YOUR TRADING COSTS TO BE? ) HOW DO YOU MEASURE YOUR RISK? WHAT LEVEL OF AVERAGE RISK DO YOU TARGET? ) DO YOU EVER USE ANY DISCRETION WHEN PLACING TRADES? ) HOW MANY CHANGES HAVE YOU MADE TO YOUR SYSTEM IN THE LAST 12 MONTHS? AUTHOR BIO S

3 Introduction Choosing a fund manager isn't easy. You can look at their track record, but few managers have been managing money for several decades; the length of time you'd usually need to determine if they are adding any alpha in a statistically significant way. You can try and ascertain whether they seem to be clever, and appear to be doing something sensible. But even a brilliant fund manager might not be able to explain how they make decisions. It is also difficult to know if their strategy will continue to work, or if they will stick to it. It should be easier to select a fund manager who uses a system to make their investing or trading decisions. You can see a back test which can go back much further than their actual track record. They can explain their system to you, even if they won't give away the finer details. Then you can judge for yourself it makes sense, and will continue to be successful. However there are plenty of poor systematic managers out there. As we will discuss below, backtests can be misleading, or unrealistic. You need more than a good back test to see if an investment is worthwhile. So here are seven questions to ask your potential systematic manager. Note: in some cases we have cheated and included compound questions. If you d like to know more about Systematic Trading you can read Robert s book, Systematic Trading: A unique new method for designing trading and investing systems (Harriman House). 3

4 1) What market and academic experience do your researchers and portfolio managers have? Designing trading systems requires two kinds of quite different skills and knowledge. Firstly it is very useful to come from a scientific background such as physics, mathematics or engineering. If you're academically trained in one of these disciplines then you are more likely to design robust automated trading systems. Also if you have been trained in statistical methods such as econometrics then you should, in theory, do a better job of fitting your trading system. However those who are scientific black belts but neophytes at trading are prone to making serious errors. Some of the biggest blow ups in trading history have been caused by extremely clever and well qualified people making silly mistakes. The meltdown of Long Term Capital Management in 1998 happened despite the fund having two Nobel Prize winners on their staff. Derivatives backed by subprime mortgages were radically overpriced before they crashed in value in 2008, thanks to traders using a clever model created by a very smart guy with a PhD. Other examples include the quant quake of summer 2007, and the losses suffered in the Swiss France devaluation of January In all these cases the rocket scientists had created a model which was a good approximation to reality most of the time, but ignored the very different dynamics of a market crisis. This is why it's also important to have people with real market experience. Experienced traders, bloodied by the market crashes of the past, are more likely to design trading systems that can cope with these extreme situations. Other common errors made by academically inclined portfolio managers include underestimating the costs of executing an order, and ignoring critical elements of market structure in back tests, such as short selling constraints. A successful systematic fund will include people with strong relevant academic backgrounds, and those with real life trading experience. Ideally the senior managers of the fund should combine both those attributes. Also beware of investing in funds which have the relevant experience, but where it is split between different teams who do not communicate with each other. 4

5 2) Explain your strategy in terms my grandmother would understand A systematic fund manager should be able to explain their strategy to you. It does not need to be revealed in its entirety, every line of code exposed, but you need to have some idea. What's more they should be able to explain the strategy in terms which are transparent and understandable. Do not accept bland platitudes like We employ high probability indicators and robust risk management. Why does a strategy need to be explained and understood? We can think of several reasons: Being able to explain something in simple terms is an indicator that you have a deep understanding of it Not being able to explain something properly is an indicator that your manager doesn't understand it, or worse still is the next Bernie Madoff (As Mr Madoff himself said: We run a split strike conversion strategy and opportunistically time our purchases, buying put options to protect our downside ) If you know what the manager is doing then you can judge if they are likely to be any good at it, based on their experience and skills If you understand the managers strategy you can make a judgement as to whether it would have really worked in the past, have an opinion on whether it will still work, and use an appropriate benchmark to compare with its future performance. Ultimately, wouldn't you like to know what your manager is doing with your money? The statement in terms my grandmother would understand' applies equally to the rest of the questions below. Do not be fobbed off with gobbledegook. 5

6 3) On average how many instruments do you have in your portfolio? How did you choose them? How did you choose the portfolio weights they have? Diversification really is the only free lunch in finance. It is much better to add diversifying assets to your portfolio, than to come up with new ways of predicting the instruments you already have. Beware of a manager who has a relatively small number of assets, or a highly concentrated portfolio. Ask them what evidence they have that this makes sense, and how they came up with the selection they have. What information did they use? Did they use a statistically robust process? If they have a strong belief that some of their instruments will outperform others in the future, then you should probably express scepticism there is rarely enough evidence for this assertion. For very large managers liquidity constraints will prevent them from giving the same allocation to small emerging markets as they would to highly liquid US counterparts. But a large manager does have the benefit of accessing a wider variety of instruments, particularly OTC markets which require an up-front investment in back office staff and technology. 6

7 4) What do you expect your trading costs to be? As a proportion of your expected returns? What were they in the last 12 months? Unless they trade at a glacial pace, good fund managers and traders are obsessed with trading costs. A manager who doesn't know what their costs are or what they have been recently should be avoided like the plague. Personally we would also be very wary of anyone who pays too much in costs, or of a manager who justifies high costs with an even higher expectation of returns. Let's consider an example with actual numbers. Take a systematic CTA targeting 15% annualised standard deviation of returns. With a very diversified portfolio it would be reasonable to expect a Sharpe Ratio of 1.0 before costs, or 15% a year (Sharpe multiplied by annualised standard deviation). Avoid paying more than one third of those returns in costs, and ideally a lot less. So the absolute maximum you should pay in costs would be 5%, which would reduce the return to 10%. So in this example be careful of an optimistic manager who expected to get 30% a year, or one who was paying 6% in costs. 7

8 5) How do you measure your risk? What level of average risk do you target? What is the minimum and maximum level of risk you'll usually have? What risks are you exposed to that your main risk measure doesn't capture? How much leverage do you use on average, and at peak levels? The only thing that fund managers should worry about more than costs is risk. There is no single right answer to any of these questions, but a manager who cannot give good detailed answers is probably a disaster waiting to happen. You will need to be comfortable with the level of risk. Be particularly wary of a manager who tells you that they have successfully hedged out all or the bulk of their risk, and they can't think of a situation in which they'd lose a lot money. This shows both a lack of understanding, and a lack of imagination. If a fund subsequently has much higher risk than expected (unexpectedly large losses or gains), then something is wrong. Leverage is a specific and very crude way, of measuring risk. But it is the only measure that will tell you what your exposure is to a total loss. Leverage varies across different asset classes. Comparing the leverage on a EuroDollar futures contract to a Chinese equity wouldn't make any sense. You'll need to ask the manager to breakdown their leverage calculations, and to separate it out for different instruments. What the figures will give you an indication of whether a manager has a strategy with low natural risk, which they have had to leverage up to get a decent return, such as a fixed income arbitrage strategy. It will also tell you if a manager is trading an instrument with very low natural risk, such as the CHF/EUR currency pair between 2011 and Eventually these low risk investments will become high risk; usually quickly and most unexpectedly. If your manager has leveraged up too much then you will suffer serious losses. 8

9 6) Do you ever use any discretion when placing trades? Some advanced systems will be able to feed orders automatically into the market, taking account of available liquidity. It is also sensible to use the discretion and experience of human execution traders to time, and to smooth, orders. Dumping a massive order into the market in a single trade just before a major economic announcement would be pretty stupid. But it should be rare that an order cannot be completed in the same day due to liquidity shortages. If this occurs frequently it suggests that the manager is taking on positions that are too large, and/ or trading them too quickly. The order that is done should always be in the same direction as the system requires (buy or sell). Otherwise you do not really have a systematic manager you have a discretionary manager who sometimes uses a system. 9

10 7) How many changes have you made to your system in the last 12 months? What were the reasons for each change? Changes to systems can happen for a number of reasons. Some of these make more sense than others. Here are some reasons, in descending order of preference: 1) Adding diversifying instruments: As we have already mentioned, we are big fans of this. 2) Adjusting portfolio allocations as market capacity changes: Also makes sense, but shouldn't be happening too frequently. 3) New models: Might make sense in a multi-strategy fund, but in a single style fund would you want to see the manager adding a new kind of trading rule? 4) Improvements to the model: Some funds think that it is a good thing to have lots of smart researchers on their payroll, and that clients, like you, will be impressed. These smart people need something to do; and making continual improvements is a way of keeping them occupied. You should judge for yourself whether you agree. Most improvements to a model are unlikely to be statistically significant. Improvements often make the system more complex, and difficult to understand. 5) Risk management: A good system should do its own risk management, cutting positions automatically when risk increases. If the fund manager keeps overriding the system because of perceived risks that the algorithm can't see, then that is a red flag. In an ideal world changing your system should be a rare event. At best you will incur extra trading costs from frequent changes; and at worse you ll significantly reduce the returns that your system could have made if left alone. 10

11 Author Bio s Robert Carver worked in the City of London for over a decade. He initially traded exotic derivative products for Barclay s investment bank, and then worked as a portfolio manager for AHL, one of the world s largest systematic hedge funds before, during and after the global financial meltdown of He was responsible for the creation of AHL s fundamental global macro strategy, and then managed the funds multi-billion dollar fixed income portfolio before retiring from the industry in Robert, who has bachelors and masters degrees in Economics, now systematically trades his own portfolios of futures and equities. Robert blogs about finance and investment at qoppac.blogspot.com. He is the author of Systematic Trading: A unique new method for designing trading and investing systems, which was published by Harriman House in September For more information see Niels Kaastrup-Larsen is a Swiss-based dad, husband, entrepreneur and hedge fund manager turned podcaster. His podcast TopTradersUnplugged.com is the leading podcast within the hedge fund industry. Niels divides his professional time between, his full time job at DUNN Capital Management, his podcast and his family s charity kidsheart.org. Niels wants to revolutionize the hedge fund industry as well as the way schools are equipped to handle cardiac arrests and other heart related emergencies following his own son s cardiac arrest in The bio could be much longer, but in the end, all you really need to know is that Niels is a father, a husband, passionate about hedge funds and CTAs and a man who cares deeply about, loves, and admires those closest to him and is humbled and grateful for the opportunity to create, to connect and to serve. 11