Saturday, 28 March 2015

Frank(furt)ly speaking

14:26 Posted by The Thalesians (@thalesians) No comments

I recently spoke at MathFinance's annual conference in Frankfurt, which I enjoyed greatly, organised by Uwe Wystrup and his team. I disagree that speaking about your trading strategies will somehow undermine what you are doing. I very much disagree! It is often possible to describe the vagaries of a strategy without going into all the details (and it is in the details where a quant adds the most value to a trading strategy). If anything, I find presentations an ideal way to solicit feedback about my approach to markets. Speakers can potentially gain a huge amount of benefit, tackling questions they never thought of!

In between the machinations of the conference, I had a few spare moments to see Frankfurt. Compared to my home city London, Frankfurt is admittedly much smaller with nearly 700,000 inhabitants (although this swells considerably during working hours). As a city it is easy to navigate on foot, given its size. There is much more to the city than the ECB, Bundesbank and the rest of banks which have their headquarters in the city.

As in the rest of Germany, the bread is very fresh and the pretzels are excellent (I think I lost count of how many I had in Frankfurt). There's a wide variety of cuisines available, including an excellent Thai restaurant nestled nearby the shopping district. Along, the river Main runs a footpath, which offers a leisurely way to stroll through the centre of the city. Perhaps one of the most unexpected of sights is the Eiserner Steg Bridge, covered in lovers locks, something you might associate more with Paris than a city like Frankfurt.

In essence, any city is a evolving complicated concoction of many people, places and experiences. Our minds love shortcuts, simple descriptions, often veering towards the negatives, rather than the positives. Cairo becomes "busy", Hong Kong becomes "polluted", Frankfurt becomes "quiet". However, this vague stereotyping of places, becomes little more than exercise in imperfect approximation. Whether or not we might agree with such labels, we often recognise them.

This is true of market. In markets, the narrative often hinges around a small number of factors (maybe even a single factor). Whether or not you agree with a current market narrative, attempting to fight it can be futile. Say we think the Fed are wrong not to raise rates quicker than they likely will end up doing. This is irrelevant from a trading perspective. We do not get a vote on the FOMC (or indeed the ECB, pictured above from the Eiserner Steg Bridge).

There is a difference between wanting some event to happen (or believing it to be the best course of action) and the most probable outcome. It is the latter which matters more when we trade. We might not like a market narrative, or an inaccurate label. However, if that narrative is the predominant one, we have to learn to recognise it and act accordingly when it comes to the market. There's little point being right about a factor the market doesn't care about at that time.

Perhaps, one way to illustrate it is as follows... in Frankfurt, it's better to have a pretzel, made locally, rather than a macaron, which is imported!

Like my writing? Have a look at my book Trading Thalesians - What the ancient world can teach us about trading today is on Palgrave Macmillan. You can order the book on Amazon. Drop me a message if you're interesting in me writing something for you or creating a systematic trading strategy for you! Please also come to our regular finance talks in London, New York and Budapest - join our Meetup.com group for more details here (Thalesians calendar below)

17 Apr - Budapest - Impact of bitcoin - Tamas Blummer & Panel featuring Izabella Kaminska / FT)
22 Apr - New York - How Smart Money Invests and Market Prices Are DeterminedLasse Pedersen
29 Apr - London  - Global macro & UK election panel - Eric Burroughs / Reuters, Mark Cudmore / Bloomberg, Jordan Rochester / Nomura, Jeremy Wilkinson-Smith / Independent & Saeed Amen / Thalesians

Saturday, 21 March 2015

Quants build bridges

15:32 Posted by The Thalesians (@thalesians) 1 comment

"I revel in the dream of numbers" is the first line of my novel, a novel which I suspect I shall never finish. I suspect if I end up writing a second book, it will most likely be on numbers and markets, a follow up to my first, Trading Thalesians! For whatever reason, I just enjoy mathematics, no matter how abstract it can somehow seem. For the most part, I suspect most people do not share such a deep love of numbers. Despite this, mathematics and its closely related fields are becoming more crucial to modern life. Mathematics builds the bridge, on which humans are increasingly progressing.

It is also true of real bridges, which have been an important factor in facilitating the expansion of economic activity. Perhaps one of the most famous bridges in the United States, is the Golden Gate bridge in San Francisco, which spans the Golden Gate strait and is 1.7 miles long, its distinctive red colour visible from a distance. If you ever get an opportunity, I would recommend walking the entire length of the bridge. The sound of the traffic is not perhaps, the most appealing aspect of the walk. The view, however, is spectacular. On one side lies the expanse of the Pacific Ocean and on the other the San Francisco Bay. Taking the time to walk along the entire duration of its length gives you a true appreciation of how much of an achievement building the bridge was in the 1930s. Of course, in order to walk along the bridge, you do not need to have any understanding of precisely how the bridge was engineered. Mathematics was used as a tool by engineers seeking to build the bridge, so that today we can cross it.

The markets are no different. Mathematics can help to explain a lot about markets, although it is very often in the background. Indeed, a similar point was made very well by Vyas from CPPIB at a talk which I heard recently at Global Derivatives USA. He stressed that he viewed quant as a tool, when it comes to investing. We can illustrate this point using a few examples.

When it comes to discretionary traders, I so often hear the comment, "No, we're just discretionary traders, we don't use any quant stuff". When we distil precisely what a trader does, it is inescapable that all traders will end up using some quant techniques to some extent.

Let us take the example of volatility traders examining volatility surfaces. They might be trying to understanding the relative value of certain options. They wants to understand which parts of the volatility surface are cheap and which are expensive. They might also look at various tenors to try to understand anomalies in time space. Let us consider how much data is needed to do this effectively. Each tenor has 5 quotes and you might have at least 10 tenors (often more). That's 50 points for every trading day. If you have 5 years of data that is 12,500 points. Across 10 currency pairs that is 125,000 points. How else are you supposed to crunch that data, without doing some heavy duty coding and mathematics to summarise it into a digestible form to make a trading decision? At the end of the process, a discretionary trader would still have a final say on the whether to buy or sell based on the output from the analysis. This contrasts with a systematic trader, who would go with whatever the trading model suggested. However, much of the thought processes of these two traders are very similar.

Whilst we have illustrated this point with options, we could easily have the same issue with many other areas of the market, notably with stocks. If a trader has a mandate to trade any of the stocks in the S&P500, how can he or she possibility know all the fundamental details of all 500 companies? Quant tools can help flag trading opportunities, which a discretionary trader can dig into in more detail, before deciding to trade. Alternatively, we can create the final output to be a buy or sell signal for a systematic trader.

So is mathematics a pancea for markets? Not really. Like any tool, it can be misused. As I discuss in my book, Trading Thalesians, mathematics should be used as a tool to explain how markets work. It should not be used to proof something which you do not know is true!

If you're a discretionary trader, don't always dismiss what quant analysis can do to help you in your decision making. As an example, it can help summarise very large datasets, so they are easier to interpret from a discretionary perspective. Very often, you might well be walking on the bridge mathematics built, but perhaps you might not realise it. Drop me a message and I might be able to tell you more! Equally, if you're a quant trader, don't dismiss what discretionary trading can teach you.. that will be a theme for a future article.

Like my writing? Have a look at my book Trading Thalesians - What the ancient world can teach us about trading today is on Palgrave Macmillan. You can order the book on Amazon. Drop me a message if you're interesting in me writing something for you or creating a systematic trading strategy for you! Please also come to our regular finance talks in London, New York and Budapest - join our Meetup.com group for more details here

Our next London Thalesians talk is on Wednesday 25th March by Matthew Dixon on speeding up high level languages, tickets here.

Sunday, 15 March 2015

Learn through travel

19:49 Posted by The Thalesians (@thalesians) No comments

Most of my travelling has revolved around work over the past few years. Travelling for work generally seems like vision, which is blurred much of the time and then suddenly sharpens for the briefest of moments. Sharp, running from meeting to meeting. Blur, staying in nondescript hotels. Sharp, e-mail a client following up from your meeting. Blur, consuming food purporting to be edible on flights. The list is endless and I am sure this seems familiar to some of you reading.

Hence, it was with some excitement that I went on an actual holiday recently to Hong Kong, although admittedly I still gave over one full day to meetings! Whilst reading about Hong Kong gave me some ideas of what I would see, it was a poor substitute for describing the experience of visiting the city. The skyline was all the more vivid in real life (see above). Indeed, the skyscrapers lights leapt from beneath the swarm of fog, in a way that a photograph struggles to capture. The bustle of the streets was something I had expected, yet it it seemed far more hectic in real life.

More broadly, the richness of the travelling experience is simply too rich to somehow condense into a few words. It would be difficult to say that reading a book about a city would provide a more vivid snapshot, compared with visiting it. A writer can try. A good writer might even succeed, yet in most cases, a book would compliment a visit rather than replacing it.

In markets, there is the obvious parallel when it comes to researching a market and having experience in trading. It reminds me of several questions I have been asked repeatedly during my decade in markets. 

First, what does Renaissance Technologies do? I have no idea, is the answer to that. Although my suspicion is that they somehow manage to run an insane number of strategies and adapt these effectively over time. I have no belief in the idea of a secret sauce in markets (and more can be read about that in my book Trading Thalesians).

Second, perhaps more pertinently, an even more common question I have been asked is, if your research in systematic trading strategies is so good, why don't you trade yourself? In my very first years in the industry this was a pertinent question. Neither my personal capital nor my firm's capital (Lehman Brothers at the time) was run on any of my systematic trading strategies. 

However, later, I was lucky enough to have capital placed on my systematic trading strategies both at Lehman Brothers and later at Nomura. Even then, the idea that I had skin in the game, to use an expression popularised by Nassim Taleb, was inaccurate. My downside from any of these strategies was limited to losing my job. Somewhat ironically, one of my best years for one of my trading strategies, was in the year that Lehman Brothers filed for bankruptcy and as a result, I lost my job in any case, when the company closed.

In the summer 2013, I quit my job as an Executive Director in FX at Nomura, primarily to publish my book Trading Thalesians and mainly to consult on systematic trading strategies for the Thalesians, a company I co-founded. At the same time, I decided it was about time to have my own skin in the game and to invest in my own trading strategies. My live strategies have had a Sharpe ratio of over 2 over the past 2 years I have run them. I can safely say it has been an great learning experience! Just as with travelling, an important way to learn about markets is to have your own capital at risk, once you have enough experience to do it (a crucial caveat). As I have said repeatedly, there is no way to backtest pain from a drawdown. Once you have a real life drawdown with your own cash, it focuses your mind in a way that no amount of research or reading can, from my personal experience. Of course, I've made mistakes. There are things I wish I had done better, notably I wish I had taken more risk in my trading, given the favourable results! There are also strategies I researched which I didn't end up trading, which would have ended up being profitable. 

I still think research and reading about markets is crucial, as is discussing the market with fellow practitioners and academics. However, I believe these should be done as a compliment to trading in the market. Does this mean that everyone should consider trading their own money? Not necessarily, in certain instances strategist are not always going to be good at developing strategies as they are at risk managing them. Furthermore, in some instances, it might be complicated by compliance issues.

In retrospect, everyone, who asked me that second question had an excellent point! At least now I have an answer, whenever anyone asks me: yes, I have traded my own trading strategies! That perhaps is the most valuable outcome of this whole exercise, gaining this experience and helping it inform my research into systematic trading strategies.

Like my writing? Have a look at my book Trading Thalesians - What the ancient world can teach us about trading today is on Palgrave Macmillan. You can order the book on Amazon. Drop me a message if you're interesting in me writing something for you or creating a systematic trading strategy for you! Please also come to our regular finance talks in London, New York and Budapest - join our Meetup.com group for more details here.

Saturday, 7 March 2015

Macarons, sandwiches & creating trading strategies

07:48 Posted by The Thalesians (@thalesians) No comments

Archaeological evidence suggests that humans have produced bread for at least thirty thousand years. Meat has been a part of humans' diets for much longer, when we were hunter gatherers. In recent millennia, animals were domesticated making it somewhat easier for humans to consume meat. However, whichever diet you choose to have, executing it requires adequate time to actually eat. When engrossed in a task which requires significant concentration, it can be easy to lose track of time, only to notice the gentle rumblings of your stomach reminding you to eat. I have been guilty of this on many occasions, in particular when I was writing my book (Trading Thalesians) and faced an impending deadline!

In the eighteenth century, for John Montagu, his mind was on gambling. He could remain at a gaming table for hours, time enough for hunger to strike. One solution would be to simply get up and have a meal elsewhere. However, this would interrupt his play. Instead, he suggested that his servants create a meal consisting of two slices of bread separated by a filing. This came to be known as the sandwich, after its inventor, the Earl of Sandwich, namely Montagu himself. Fast forward over a hundred years and the concept of the sandwich evolved into the burger supposedly at Louis Lunch establishment in 1900 (although Wikipedia lists several other potential inventors). Despite ingredients for sandwiches being easily accessible for millennia, it clearly took a long time for a stroke of brilliance to combine them.

If we think of other more complicated dishes, it is far less obvious how to create them, even if the ingredients are widely available. Take for example macarons. First, whilst, they are becoming more familiar in high street bakeries, I suspect most people are not fully aware of all the ingredients you would use for them, without consulting a recipe book (unlike sandwiches). Second, the steps required to bake them are numerous and often not intuitive. Even if you follow the recipe precisely, it is very easy to make subtle mistakes in their preparation, which render a macaron looking like an unrecognisable mess, even if it does taste fine (as an aside, the red macarons in the photo, were ones I made... my many previous attempts looked absolutely awful).

When it comes to thinking up of ideas for profitable trading strategies, we can use the analogy of the creation of sandwiches and macarons to illustrate the general thought processes. We can think of "sandwich" strategies as requiring little in the way of complexity. What they do require however, is for you to think of the concept to begin with. Indeed, it might require a significant amount of work to come up with the initial idea, but the actual resulting trading strategy is relatively simple. How do you increase the chances of coming up with a "sandwich" strategy?

  • Observing the market on a regular basis, can help uncover strategies. Are there any patterns you can spot? What are the explanations for these patterns? If we look at the data are our initial ideas confirmed?
  • Around market events, what type of price action do you tend to notice?
  • When do you tend to lose money and when do you tend to make money, when you trade? Trading with real cash is like a "special" backtest which you can never forget!
  • Simply talking to others in the market, you might pick up inspiration.
  • Markets interact with the real world, you might get ideas from other areas!
  • A lot of the times, it might involve trying to mimic investor behaviour you have observed and modelling that. If you have an idea to begin with, it becomes easier to model, rather than searching in the dark.
  • As with everything, experience is very valuable for uncovering systematic trading ideas!

Macaron style strategies are somewhat more difficult to create. Here the general concept can sometimes be relatively "simple", for example, using news data to trade markets. However, actually implementing the idea in practice, requires a lot of detailed work and steps to get it right. All the points important for a "sandwich" strategy are still relevant. However, we might have other issues to deal with, if we're trying to come up with a macaron-like strategy:

  • Even before starting our backtest, we note these ideas require a decent technical base to implement, whether it's certain mathematical techniques or coding.
  • There might also be a significant amount of data which needs to be managed effectively.
  • In other words, we probably can't use Excel to implement our trading strategy!
  • Given the number of steps a final algorithm might end up having, we need to be even more careful about data mining. In particular, is there always a rationale about a particular step you using? Making a trading strategy unnecessarily complicated does not make it run better in real life.
  • Try not to lose sight of the original idea amid all the detailed implementation.
  • There might be multiple quite different ways to implement the same idea.

Of course none of this is easy, but with a bit of luck and experience it is possible to build profitable trading strategies. Even once you've created a good trading strategy, there is also the important question of how you manage risk around it, when actually trading it. Fail do that properly, and you can render even the most robust of tradings strategies unprofitable! One question remains, which do you prefer, burgers or macarons? I would suggest both for a balanced trading book (and diet).

Like my writing? Have a look at my book Trading Thalesians - What the ancient world can teach us about trading today is on Palgrave Macmillan. You can order the book on Amazon. Drop me a message if you're interesting in me writing something for you or creating a systematic trading strategy for you!