Trading and "behavioral finance"

Sort of connected to my Philosophy of Economics exam, I have been reading a long book on behavioral finance, Hersh Sherwin's Beyond Greed and Fear and in particular, I have been trying to assess whether I am guilty of the types of biases to which this book refers. Intra-day trading is very demanding psychologically and one's mood can swing tremendously during the day - yet certain "big picture" matters have to be kept in mind. It would help to know if I was biased as well.

For instance, there are days like today when the morning consists of a series of market moves that are virtually guaranteed to slowly wear you down psychologically. Whenever I bought, the market would reverse, often within literally 10 seconds of my trade. When I sold, it would rally instantly. Therefore I incurred a long sequence of losing trades - something like 25 such trades out of 30 by 3:00p.m. No major losses, just the steady 5 - 10 points per trade (I don't suffer from the sort of "loss aversion" highlighted by Sherwin)

But finally all the various systems lined up and put me long the market and, finally, the market inched up slowly rather than reverse instantly. From around 5675, the FTSE100 inched up to 5700 and just beyond. This was enough in 45 minutes to wipe out much of the days loss. Soon after, the system began to close out the positions and the system trading the US markets went short. So a move of about 30 points is the sort of thing that I am typically after.

While I am trading I don't follow any market news but apparently Greece was being downgraded to junk and Goldmans was being interviewed by the US Senate committee. As a result of something perhaps related to this, both the UK and US markets began to fall and subsequently dropped about 100 points each in the space of about 30 minutes. This sort of move is, of course, quite rare and swamps the earlier losses and gains.

The interesting thing in connection with behaviour finance is that I tend to set up my trading systems with quite small price ranges (so I can see the detail fully). But this means that the 100 point fall is way above the sort of move I am looking at. Yet I need to keep in mind that such moves can occur. In particular, to not get too dismayed about sequences of small losses when something dramatic could happen at any time - that day, the next day, whenever.

In fact, the lessons of today will bear much thinking about.

I have also been watching the Senate hearings re Goldman - this has also been rather interesting. I am appalled by the ignorance of the Senators about things I certainly consider perfectly normal within finance (after a career of well over 20 years in the markets). But that said, the way Goldmans choose to reply was really odd - and very silly in my view.

Turbulence and Starlings as a model for market trading

Over the last couple of months I have read perhaps a dozen books on trading -one or two new ones but mainly from my own collection. Years ago I spent several months developing a whole series of extremely short term trading systems, mainly based on some sort of mean reversion technique i.e. to buy when a market has been weak but is beginning to strengthen, and sell when a market has been strong and starts to weaken. But these systems take a huge amount of trading infrastructure to work and I eventually abandoned the attempts.

But catching up on what has been happening in "advanced" market trading, it is clear that this would have been the way to go. It is actually slightly disappointing to discover that I was, in some respects, on exactly the right lines, even if I wasn't able to take it forward. But it also gives me support for my current plans.

Next week I intend to have a last "practice" week and then go live with all systems the week after. As a heuristic technique for developing my ideas I have been considering how I would explain my techniques to a third party - something I wasn't very good at when I worked in fund management. My basic metaphors are "order within apparent trubulence" - how, for short periods of time, chaos disappears and a tradable, directional bias develops. I have been re-reading some very technical papers on the broad field of "complexity" and there is little doubt that there is strong theoretical support for the views I have

And pictures of turbulence have got better and better!


Lots of borderlines between chaos and order

I remember the first time I saw one of the detailed Voyager pictures of the clouds of Jupiter. Yet within the apparent chaos, the great red spot maintains a coherant unity (and has done so for at least the last hundred years). The animations of Jupiters clouds are wonderful

A different pattern of "order out of chaos" is shown by flocks of starlings in flight. There have been loads of automata models of birds in flight, it being relatively easy to generate complex flight patterns from some very simple rules. Some of these rules come under the broad heading of "mean reversion", with outer birds moving back towards the average position - this is what produces the waves that appear within the flocks as each bird adopts something like the same behavior.

There are some superb photos of starling flocks - if the first one were available as a large print I would be very happy to hang it on my wall!

Perhaps I could create my own 4 x 4 picture from other photos such as those below?

Final tweaks to trading ideas before going properly live

I am making the final little changes to my selected trading systems - the main change has been to include the US markets in the overall model, with a current weighting of UK:US of about 60:40. The UK market is tremendously international (there are quite a number of stocks within the FT 100 that I know nothing about). And perhaps surprisingly, the two aren't locked together during the day. So there will be seven systems applied to two markets (rather than the four applied to one that I was looking at this time last week)

In the meantime, a new PC has arrived. The laptop (on which I am writing this) is simply not up to running the programmes needed. In the couple of hours it takes me to set up the PC, I would have missed a really good 60pt move - just shows you that you can't take your eyes off the markets during the trading day. The new PC is lovely - especially the 24" screen. I now have to learn my seventh or eighth operating system - Windows 7.

The other main development today has been the development of a bookkeeping regime - mainly focused on the intra-day trading positions. Not a bad first effort, but will probably go through dozens of changes (as my spreadsheets always seem to do) - not much MSc work going on at the moment I have to admit

I have finished the book on the Paulson hedge funds - Zuckerman's The Greatest Trade Ever. In the light of the current Goldman Sachs case, this book is really clear on the inherent huge risk being taken by those who expected the market to fall. I think that Goldman is the victim of political events and will be fully vindicated - this looks like a prime example of hindsight bias, about which I have been reading in the last week or so

Next book - Emanuel Derman's My Life as a Quant. I already have a copy of this somewhere - no idea where so I had to buy a new copy. Always annoying when that happens, but it seems to happen with increasing regularity

Emma is 21 today!!!

A long and stressful day . . . .

Working from 5:30a.m. to 9:00p.m. today - a long and really tiring day

First up, trading. My refreshing of past statistical data suggests that a typical week is made up of one really good day, one medium day, two small wins and one losing day (which contras out the two low wins and part of the medium win). Monday was an example of a medium day. Today was a losing day. Of course some weeks might, by chance alone, have two losing days, while others might have none. I just have to wait and see and be able to cope with either.

I was very aware when I started trading again that I would need to settle back into the correct psychological rythmn - the detached view in which each day is the same in that it consists of the following of the chosen system, that the actual day's results will take care of themselves, and so on. But I am not quite there yet. So each morning I remind myself that today could be the losing day and I need to just cope with it. Yet each trade starts with the hope that it might go well, and each small loss leaves me inevitably feeling a little deflated.

That said, I have been doing a lot of work in the last few weeks on recognising what sort of trading day is occuring. Interesting I was able to come up with a detailed specification of what a losing day looks like - linked to things like points differences between peaks and troughs and the times between such peaks and troughs. Certain combinations of these just won't be profitable. This suggested a form of intra-day stop in which the systems "recognise" that the day isn't going well and call a halt to things. So today followed that methodology. And around 3:30, the systems concluded that today was not to be their day and I withdrew "to fight another day". Soon afterwards (almost immediately in fact) the market set off in a fast decline dropping more than 30 pts in 30 minutes. Such a move could have cleared the day back to breakeven. But then it rallied strongly in the next 30 minutes. Sharp "V" shaped markets are horrible for my trading approach and would have left me with a substantially greater loss. So perhaps the intra-day stop has got off to a good start. I certainly felt better at 4:30, when the V was over, than I had a 3:45, when I would have been at breakeven for the day.

But this also highlights one of the problems I am still not clear of - staying detached while watching the markets throughout the day is very tough. It is so easy to start watching every little move, hoping this trade will be profitable. I do need to step back a little.

Of course I also spent the day working on Philosophy of Economics. The morning was mainly spent on compiling revision notes from what I have been reading over the past couple of weeks. This is going quite well and I do feel that bits are beginning to drop into place. But the afternoon was spent on new reading and this was very tough today. My chosen subject was the positive / normative distinction and I have been ploughing through several long papers. I was never very interested in theories of ethics when I have studied philosophy in the past, and that is definitely brought home by this current reading.

By 5:00 or so I have had enough and spend the next three or four hours reading a trading related book - Didier Sornette's Why Stock Markets Crash - and a book about the Pixies rock band. By 9:00 I was extremely tired. Maybe tomorrow I will have a new plan for the day's work - perhaps slow down on the harder stuff so I don't burn out on it.

First day of a planned week's trading

I have set aside this week with virtually no distractions so that I might do some PH413 work ready for my exam in June and also do a full week's real trading (though at a small size as I am still bedding down systems, etc)

Overnight the FTSE has pitched higher, probably off the back of the latest Greek baleout scheme. Two of the systems gave sells during the night, with the last one giving a sell around 8:10, at which point I put on a half sized opening position. In the next hour, the market trundled down about 20 points and gave some ok exits. From then on it more or less traded sideways giving a couple of tiny profitable trade and two small losses as the afternoon dragged on. +£150 for the day, an average points day according to my research. At full size, per my planned May start date, today would have made about +£380.

Meanwhile I was able to produce about a dozen pages of revision notes for PH413 - which of course are not really revision as I hadn't done the work in the first place. But some ideas are beginning to set in my mind. I am clearly not up to taking the exam yet though!

Briony Campbell's "dad" project

The Saturday Guardian featured an article on the photographic project by Briony Campbell of her father's fight against cancer. No doubt like all the other people who have lost their father to cancer, I found the feature very moving. It is about this time four years ago that my father was going through the same process. He had just a few months left to live, his weight was dropping rapidly, he couldn't eat much, he was tired all the time . . . . and so on.

It didn't occur to me to take photos of my father at that time, but then Briony Campbell was doing an MA in photographic studies at the time, so no doubt it was a more normal part of their day-to-day life. Strange how the passing of time since 2006 hasn't really made things much easier.

The "full project" is at www.brionycampbell.com

This last picture is especially moving I thought

Emma's best pictures from New York

Emma has a new camera and was busy testing it in New York


Daughter and I having breakfast at Le Pain Quotidien, Bryant Park

Wife and myself in Times Square - the classic cliche picture!

Wife and daughter at The Cheesecake Factory, Long Island

At the Bronx zoo

Lemurs

Getting ready for our helicopter ride

The old and new Yankee Statiums (statia)

The Upper West Side - somewhere in there is a bookshop with first editions of Kepler's Rudolphine Tables and De Stella nova!

Looking north from the observation deck on the Rockafeller building

Emma and the Empire State

Looking south

Current thoughts on trading & Penetration live

The statistical testing I'm currently doing is progressing well. I am of course building on huge amounts of relevant data that I have still from 2000 to 2006. I have used this data to produce many families of applicable models and have selected a number of "representative" models to "blind test" on current data. The results from this are very encouraging. Broadly speaking, the models current performance is pretty much identical to that of its previous testing period. This gives good support for my current trading plans.

I have also been reading lots of relevant market related books published over the last few years (when I have not been paying attention to the markets at all). One of the newest books published in my field has been Scott Patterson's The Quants. This was the book I bought and read during the recent New York trip. While not having much detail, there were one or two clues that have supported me in my testing. For instance, Renaissance Technologies, no doubt the finest "Quant" fund, was described as aiming for profitability not just monthly, but weekly, daily, even hourly. Descriptions of other funds also suggested this same type of focus. But to achieve this requires that the trading systems being used be extremely short-term and probably linked to profit target exits rather than system reversals - in other words, you look for a particular "set up" on intra-day data, jump when it occurs, and exit on a fixed gain.

When visiting mum the other day, I brought back some boxes of trading books and magazines and have been skimming through these the past few days. One long interview that I have studied in detail was with Doyne Farmer from 2000, just after he had left the Prediction Company, documented so well in one of my favourite trading books, Thomas Bass's The Predictors (which I have also re-read recently). This interview also gave a number of similar clues to how their systems worked in the manner outlined above. Years ago I visited a hedge fund manager as UBS and he showed me the Prediction Company's "black box" chugging away in a corner, which he claimed had a Sharpe ratio of 16. Again this is only possible with very short term, "in-and-out" systems.

So I have worked through my entire collection of magazines such as Futures, Technical Analysis of Stocks and Commodities, Active Trader, Absolute Return, Eurohedge, etc. In the latter, it is really amazing how many names I came across that I had sort of put out of my mind - funds like Mulvaney, Vega, Rubicon, IKOS, Weavering, etc. It is also strange reading them in one go to find just how many dropped out over the period. I have little regret about starting my hedge fund though it didn't succeed. So few do, and those that don't have some pretty clever people involved - there really is no shame in failure in this most-competitive field.

Other fascinating reading has been interviews with Nassim Taleb (before Fooled by Randomness) and Victor Neiderhoffer (post first blow up, pre-second). I am also studying Andrew Lo's Hedge Funds: An Analytical Perspective in great detail.

But overall, things are going well. The green light is set for late April / early May when I will combine trading with revision for my exams

And finally, I have acquired another recent recording of Penetration from their recent tour as well as two shows from 1977 and 1978. Superb stuff! I also found an old video featuring live performances by punk bands in the late 70s, one being Penetration at the Electric Circus in 1977 doing "Don't dictate". I feel quite nostalgic - and also very clear that modern bands have little of the intensity of punk.

The Boat Race

The Boat Race always sort of lurked in the background for us and it is only in the last couple of years that we have paid it closer attention, entirely because Emma's boyfriend is a Cambridge rower. For much of the past months he had been included in the main boat but at the last minute was bumped down to the Goldie reserve boat, even after the press announcements has occured. Such a shame for Code, who had invited several friends from home in the States to see the race - in addition to his mum and dad who had travelled over to see the race again

Emma had been to London a couple of day's ago to see one of the practice runs and was on an early train today, meeting various people near the start line.

And we were tuned in mid-afternoon to watch the build up, during which there were various clips of the Cambridge crew training which featured brief glimpses of Code. Then the news came in (incredibly briefly) that Goldie had won the reserves race, so at least Code got that. Again a brief clip of them crossing the finishing line and collapsing as they do

Then the first boat race and, much to many people's surprise, Cambridge won again, holding off the Oxford boat on the bend where Oxford had the advantage and pulling away when the bend favoured them. All very exciting stuff and just what they all wanted to ensure a full night of celebrating in London

Emma appeared at Didcot station mid-evening Sunday, rather tired as you would expect. They had got to bed about 4:00am, probably about the time I was getting up to start the day's study!

Maybe Code will stay on at Cambridge for one more year and will get another shot at the first boat - such a shame he wasn't in it this year

Last biography class

I missed the penultimate biography class last week and tonight was the actual last one. The last two sessions are taken up with us reading from our own work. I didn't have anything to read and only got started on something this afternoon around 2:00p.m. My chosen writing was a selection of episodes from Kepler's life broadly drawn from the theme of religious confessionalisation. It ended up showing that such a topic will clearly be one that is scattered among a full biography rather than gathered together into a single topic. But I was quite happy with what I'd put together and the feedback was ok. Nicki in particular had some good points, but then her background is more academic than the others.

Only two others did a reading today - indeed virtually all the people who read last week failed to turn up tonight. But Katie did do a reading from her proposed biography of a Tibetan Lama who she knows well. This seems a well developed project already. Katie mentioned that she had been to Tibet several times and was thinking also of writing a travel book. I would be very pleased to read both at some point.

We had time for a quick drink afterwards. This course has been quite surprising to me - much more academically challenging than I had expected, and all the better for this. Once again I am left wondering whether any of us will keep in touch. Our email addresses have been swapped, so who knows.