What a day...

September only started and has been a rollercoaster already. As market sentiment kept going down, profits kept coming in as participants kept flowing into "safe havens" such as the yen and swiss franc. Well, then the market reversed (as it always does I guess...) and with the better than expected US labor market data an outflow of capital out of those havens was triggered. My short position in CAD/CHF actually was hit hardest with a loss of 1.73% in a single day... The bullish piercing pattern (explanation see http://www.swing-trade-stocks.com/candlestick-patterns.html) might have been interpreted as a sign but then again my strategy is based on trends and not candlestick patterns. And so far the trend is still my friend...

The long positions in the yen are still not really moving in any way. Using a candlestick pattern again, there is indication of a trend reversal however. Today's candle might be a reversed hammer (explanantion see http://stockcharts.com/school/doku.php?id=chart_school:chart_analysis:introduction_to_candlesticks, scroll  down a bit)

In a reversed hammer bulls try to takeover but cannot do it just yet. This friday was marked by a good US labor market but negative ISM-service sentiment. This does fit in the general bullish attidue (long top) but still some holding back due to traders being slightly unsecure. This pattern needs confirmation on monday. If the chart rallies, both yen positions will be a loss as well.

Those big swings in equity are quite an experience to me as I did not quite realize how quickly those things can happen. I am definitely trying to detach my emotions from each position and each market movement but I have caught myself "baby sitting" positions, meaning to check the markets every hour instead of once or twice a day. Obviously will have to work on that but I hope my learning curve will be steep!

Now just some bookkeeping issues. I started to record monthly AND weekly equity levels to gain some performance insight. So in a sense the experiment part 2, currencies started on september 1st! Let's see where it goes. I am definitely planning on going all the way till next summer and hopefully another year while doing my Masters.

Break Out Analysis

Now that I talked about money management a bit I would like to talk about the actual signal generation again. When I started with this whole experiment, I was using new high and lows as entry points. The rationale behind this being that once a new high is formed the buyer side (demand) finally increases above a certain level, triggering a momentum move upwards. Vice versa when looking at a new low. Now, I am using the same principle but more of a graphical execution. I look at the daily chart of the past 6 months and identify strong supports or resistence areas in the chart. Once those are broken, a signal is triggered. This is called a break out system (see http://www.investopedia.com/terms/b/breakouttrader.asp). Again the rationale behind it is that once such a level is broken, there is a fundamental change in the supply and demand forces of the underlying asset. For example when a resistence is broken, there are now more people willing to pay a higher price and less people are willing to sell at this price than it was the case for the time period before. This can be triggered through events such as economic news but also a shift in the general attitude of traders. Because it is a relatively efficient and quick way to identify entry levels, a lot of markets can be covered. Also Since (most the time) a former resistence becomes a new support level, it is easy to place stop-losses just below/ above those levels. Sounds good doesn't it?! There is a problem however: It is not an exact science to identify those areas. Only experience and practice improve results, and even then there is no quarantee that it might be a false break out...

For now I am using 2 particular situations:

1) Break out of a price range. by looking at the graph it becomes apparent what I mean. Highs and lows are around the same level respectively for a while. Thus they are forming resistence and support levels. Soon or later the market WILL START TRENDING and break one of teh levels in teh process. By placing an entry order just above or below those levels gives a nice opportunity to enter a trend early.

This graph of the USD/CAD shows a real life example. The USD remained within a certain range for almost 6 months.Both support and resistence levels are pretty obvious. Then in early August teh USD gained in value and the resistence is broken on a closing basis. A clear entry signal for a long position which would have performed nicely indeed...

2) Old high/ low trend break. When a trend is forming old highs are lower than new highs and old lows are lower than new ones in a uptrend. Vice versa in a down trend. When talking about a break out in this context, it means the asset price is overtaking the previous high (vice versa on low). At this point a trend is established and it is time to move in.

Here is another real life example. Strictly speaking There should have been an entry after the support break out near the top, but a trend as such is only established at the second green circle. This second break out is where the position is added to and the stop-loss moved (at the latest) or the resulting trend line used as the stop.

Alright this short overview should explain when I enter positions and why (demand-supply shift, trader sentiment shift). As such many people utilize a break out system in one way or another. For moreinformation a google search should help...

Yen Rise

China is overtaking Japan as the second biggest economy today, the Japanese GDP is stagnating, unemployment rising... In other words Japan is continuing its declline since the 90s. But why is the Yen at a 11 and 9 year high compared to the USD and EUR respectively?

1) Carry Trade: Borrowing in Yen at low interest and investing in a higher yielding currency thereby increasing demand? Has been going on for years plus the USD shows no interest either. So why would that be the reason? Doubtful...

2) Psychology: The Yen as substitute for the USD in tough times? Doesn't go well with the economic data plus there are  better subs such as the Swiss Franken in my opinion.

3) Technical trend? Definitely a trend there but such a strong trend usually has underlying fundamentals. Again not exactly sure which ones. Maybe the unattractiveness of US bond yields and other safe haven yields.Most likely it's some combination of all of the above.

In any case an entry order for USD/JPY and HKD/JPY for a short position is entered. The trend might be ending (end of 3rd Elliot Wave?) but then again it might not. As Dennis would say: Buy high, sell low...

Short update

Alright, learning how to program does take longer than expected, especially when I'm also learning Romanian, working full time and preparing a dissertation... Therefore I used some of my old data and a few new trials to semi-test a simple 3-month break out system on currencies. Performance is decent but should be pretty good using the proper risk management system. After a bit more thorough research in that area, I laid out the following:

• Initial risk per trade: 2% of free equity
• Max risk per trade before reducing position size: 3% of free equity
• No more than 22.5% of capital at risk at any time
• No more than 2 positions in 1 direction per currency (e.g. no more than 2 positions betting on a stronger euro, however opposite positions do cancel each other)
• When adjusting the stop-loss AND simultaneously favorable currency movement, lots might be added to increase exposure to 3% of free equity again
• Obviously number of contracts will always be rounded down
• positions will be adjusted daily at a set time once US markets have quiet down, so it should be early night in Europe, but I will see how well that works

I hope those rules should keep me from overexpusore and overtrading. The 22.5% of capital at risk might be a bit high as it is more than 80% of the Kelly formula's suggestion but at the same time gives enough room to hold varies positions. The no more than 2 position per direction rule should prevent any overexposure to any specific currency. I did look at some other ways to tackle this, for example geographic region, interest rate setting central bank or major industry types of the economy but in the end this solution seems to be the easiest to implement.  Maybe the rules seem a bit confusing and or weird right now,but I think they do provide a good framework to start with (again) and I will bring them to live with some cool trades :)

I wish I could try this system on futures but my account is obviously not big enough to  do so in a manner that money management actually exists (trading only 1 contract each timeis no good money management...) and stocks have astronomical trasnaction cost. Therefore I will soley focus on FX microlots (1k lot size) for now and hopefully will be able to move up  to standard lot and maybe CFDs (so I get stock and commodity exposure) next year.

The account is set up, the money on its way and I should be starting to trade again by the end of this week! Until then...

Still alive

So, the experiment is indeed still alive. I abolished live trading for a little bit right now, because I did some back testing once more and was not too happy with the results. When I was optimizing protfolio performance disregarding any transaction cost and then applying the cost afterwards, I got good results for the theoretical performance but transaction cost ate away all the profits... On the other hand when I tried to lower cost, theoretical performance was terrible. As a result I decided to look into FOREX trading as cost might be more favorable there. Only problem is that in order to test the system I do need to learn how to program in Metatrader. This will take probably until the end of summer as I'm working in my internship as well as I'm trying to learn Romanian at the same time. I will give periodical updates on my progress and plan to be back on live trading once I start final year at university :)

Meanwhile I will try and follow the markets to see if the general idea works in indecisive market times such as the current one...

Natural Gas Exit and March Performance

On Easter monday natural gas finally hit its stopp loss. I only noticed today as I was not online on monday but that should not be a big problem since the price at Henry dropped a bit again today. Of course this could mean natural gas' intense upswing on monday might be a technical reaction. In any way I exited the position today with some 14% profit after fees. The hypothetical profit according to the system is some 17%, which is still within my expected deviation of up to 5% due to fees, timing issues and tracking error of the financial instruments used. Here is the final graph:

Sugar, nickel and most indices are maxed out positions that are still going up. I will provide uptdates on those as they come but I hope the upward trend will remain in tact, which is to be expected regarding the FEDs bubble building low interest rate politic...

As a final note I would like to mention that the portfolio made a return of 1.75% during March. Performance could have been a bit better but profit taking by market participants during the last few days of the quarter shed away something like half a percent.

That is it for today. Going to celebrate the internship offer I received today for the summer now and will be back soon with some more news and thoughts :)

Updates on Cotton, Sugar and Natural Gas

As the major stock markets still seem rather indecisive about their further direction this year, commodities are the place to go for any trends so far. Yes, most of my trades have been unprofitable, but there are two positions that promise to be big home runs and pay for the other losers: sugar and natural gas. Before I look  at those two, let me sum up the cotton trade from the last post.  


Cotton long trade, 02/16/2010 to 03/10/2010

The chart shows some profit but since there are two points where I added to the position, the overall profit was around zero after transaction cost. Still, nice little trend that made me believe more in my system.
Now, let's look at the two most promising trades in my portfolio. I like to point out that both were entered solely due to technical signals and that no fundamental analysis of any kind was invloved (I'm neither an expert on natural gas nor sugar). Nevertheless, both sugar's and natural gas' declining trends, made me curious and I tried to find some sort of fundamental reasoning. Here we go: 

Sugar short trade, entered 02/12/2010

Sugar hit a 29-year high at almost 30 cents earlier this year but is down more than 40% from that today. In fact today marks a new closing low at 17.63 cents (not in the chart yet). My system has been able to capture the bulk of the downward trend as seen in  the graph. Due to 2 increases in the position along the way, the performance is only around 24% instead of the 33% from the original selling signal however. Nevertheless very nice money making trend... But what is the reason for sugar's decline? After a quick google search, there seems to be an increase in sugar supply as Brazil, India and Australia report higher than expected sugar crop yields this year. Especially the predicted 12.7% increase in the output of Brazil (the biggest sugar producer of the world) adds downward pressure on the price. On the other hand China and Thailand have experienced lower than expected harvests (see http://www.ft.com/cms/s/0/51f828de-31ba-11df-9ef5-00144feabdc0.html). If sugar continues to fall and stays below the support around 18 cents, speculators might liquidate their massive long positions which would give further momentum to the collapse of sugar (finanzen.net). Obviously there is always the possibility of the price hitting a floor backed by an increase in demand (Egypt, Mexico and Indonesia are supposed to buy their sugar soon). Stopp-loss is set at around 19.1 cents, so whatever the move will be, I'm prepared...
The other big winner so far is a short position in natural gas. Natural gas is down 30% from its high in December and once more my system has been able to capture a large proportion of this trend so far.

Natural gas short trade, entered 01/28/2010

From a fundamental point of view there a downward pressing forces on both demand and supply side. "The Economist" is featuring 2 articles on natural gas in the March 13th-19th edition which explain some of the reasons. First, the development of new technology has decreased the cost of drilling gas from shale. In fact it has turned some areas in Canda and the US from importers to exporters. Similiar projects are under way in Europe and China. Especially China is keen to diversify its gas supply by using shale gas from China. In a nutshell, drilling gas is cheap and the fields in the US are massive. The demand side offers a more mixed picture. Short term consumption has fallen due to the rescession, the end of the winter in the US and environmental restrictions in Europe. Still long term demand could pick up considering electricity from natural gas is much cleaner than from today's predominant source coal. Production could be easily switched in favor of gas (the US already has massive capacities ready, "The Economist"). Therefore the historically volatile gas price may remain volatile for some time...
There are some more interesting positions in my portfolio right now,  such as the FTSE 100, Nifty 10, Brent Crude Oil and Nickel. All of their fortunes seem to depend on the rather vague future direction of stock markets though, so I decided to wait with their coverage for a bit longer. We shall see mai tirziu...
(yes, I'm trying to pick up Romanian)

Up and down and up again?

So far this year hasn't been exactly amazing trendwise. There has been some smaller trends but they weren't big enough to make money out of them for my system. Maybe I will work on a tighter system with more leverage to capture those, but that will be a summer activity... Anyways, most stock markets are hitting new heights right now, so I'm already in or in the process of getting in long, hoping for some more sustainable trends this time. We shall see. In the meantime I want to look at 3 commodity positions I'm holding:

1st: Cotton long, entered 02/16/2010, add to position 02/22/2010 and 03/01/2010

As you can see there has been a nice little upward trend in the price of cotton. In fact it has been hitting a 52-week high at around 83.5 cents. This trade shows nicely how the system plows back profits by adding to winning trades at certain new heights. The downside is that after the trend has lost its momentum, the percentage gain is marginal after adding to the position twice.


2nd Sugar short, entered 02/12/2010, add to position 02/23/2010

Again, nice downward trend from almost 28 cents to around 21 cents (almost 25%) up until today. The second and final position increase should be done today or tomorrow, if the trend holds. The next few days will be interesting, as it seems that there are quite a few supports around the 20 cent mark. If that level breaks, the sugar trade has the potential to be not only  big but hugely profitable...

3rd Natural Gas short, entered 01/28/2010, add to position 03/04/2010

Now, the Natural Gas trade is very interesting in my opinion because it shows that sometimes it takes time before the trend takes of. The entry signal was on the 28th of January and it seemed like the market was moving against that position but from the 19th of February onwards the downward pressure finally kicked in. So far the trend seems intact (add to position on the 4th of March) but it could be a downturn fuelled by speculation as a COT report suggests. So far there is no sign of an upward trend though, so I happily watch my profits rise...

Lastly a note on the equity of the whole portfolio: February was a bad month with a drawback of 2.2%. This doesn't sound that bad but please keep in mind I cannot trade any leveraged positions yet, due to restrictions from my broker... Anyways, let's see if there are a few more nice trends in March and April to gain back some  of that equity.

Trade Universe

Today I want to talk a bit about choosing the markets to trade in. When I first started I admittedly just looked at which structured products are available for which market and then included those markets in my investable universe. That's not a good approach as I quickly realized. When markets plummed in late January, my portfolio went through the bottom. In my quest to diversify, I just forgot about one important aspect: correlation.
So after getting out of many of my positions I took a closer look at my portfolio and decided to reduce the number of markets that I trade. This has the nice side effect that my transaction costs relative to each trade actually decrease but again that's not the main point. My main goal was to decrease the correlation between my markets. I did a very basic calculation of computing the correlation between each market based on daily returns for the past year. Afterwards I started adding markets based on their correlation to the portfolio (I started with mature markets, which admittedly is arbitrary again). As a result I reduced the number of markets that I monitor to 29:

Mature markets: TecDax, FTSE 100, S&P 500, Nikkei 225
Emerging Markets: Nifty 10 (India), MSCI Turkey, MSCI Greece, South Africa, FTSE Vietnam
Metals: Gold, Silver, Platin, Copper, Nickel, Aluminum, Zinc, Lead
Agriculture: Corn, Cotton, Coffee, Cacao, Sugar, Soy Beans
Energy: Natural Gas, Brent Crude Oil
Bonds: Treasury 7-10, Germany10+, Iboxx10+, Euro Corporate

A word about January's performance: BAD... I am confident that the system works, but so far only the MSCI Turkey and sugar future trades are in the money a good bit. Oil has been swinging in between 72 and 80 dollars creating a few wrong signals. Same with the Metals. Let's see if February will be better. My portfolio is definitely set for a bearish month, with short positions in gold, silver, copper, cotton, FTSE 100 and S&P500. There would be a few more shorts but I couldn't find any products that I am allowed to trade for those, which is quite annoying because the soy bean trade my system indicated would have made a nice unlevered profit of about 7.2% minus transaction cost. I will definitely have to look into those restrictions once I have some time after midterms, job applications and my job at the 2010 Winter Olympics...

Risk and Money Management

So, after entering my first trades and realizing my first losses (more on that later) I found a bit of a flaw in my money management system. Strangely all positions were supposed to be the same size, even though they clearly exhibited different amounts of risk. Just to be on the same page, by risk I do not mean the convential notion of risk in terms of standard deviation or Beta but merely the possible loss in the trade. In other words the difference between entering price and exit price marked by the stop loss . After looking at my programming in excel it became clear I should lower the risk each trade represents or otherwise I might be out of money sooner than later... Now the total amount at risk is no more than 10% of total capital. Usually it's less because this number is only achieved when I enter 36 positions (42 possible positions in my investment horizon of which 6 have a negative correlation with the other 36 and hence it is unlikely they will all create a buy signal). Also I included a buffer of 5% added risk to each trade to account for tracking error from slippage cost such as transactions, time delay or tracking error between the underlying and the financial derrvative I'm using. My first trades indicate that the 5% are quite a good proxy but I will make sure to adjust this number once I have more trades in. After applying those money management thoughts on my back test, I actually increased its annual return by almost 1% over the 10 year period I was looking at. In conclusion I now feel more confident I will be saving money to go after the big hit, which should be coming eventually.
That being sad let's look at the (not so amazing) trades since my last post. On the 13th I exited my Nickel Future trade when it hit the trailing stop with a 9% (all gains/losses include transaction cost etc.) loss.

 

Similiarly my trades in corn (-13.7%), lead (-7.5%) and zinc (-9.9%) got stopped in the same way.

Another picture perfect failing signal was demonstrated by my Brent Crude Oil trade (-9.8%). After breaking a high on the 19th, it looked like a nice upward trend was developing with the backing of rising stock markets and continuing positive market sentiment. However, teh oil price was soon to drop, hit the trailing stop and crush my hope for a medium term upward trend. In fact it seems liek markets in general will follow the oil example...

Especially last week hasn't been kind to my positions as markets seem to swing towards the bear side, at least for now. I'm starting to feel how I am still emotionally attached to some positions and how I have been wanting to take my gains on some older positions (been holding the Nifty 10 for quite a while now). This weekend should be my first little challenge as I will most likely have to realize a bunch of losses in most of my positions, even though I would rather "sit-and-hold" with "my dear" financial instruments. Alright, guess I still have some way to go until I reach teh final point of the "game" mindset. Hopefully on the right track though...

Happy New Year 2010 and here we go...

New year, new decade, time for some hands on financial experiment. After reading Micheal Covel's "Trend Following" I decided to try it myself. I mean the results of trend followers seem rather impressive compared to traditional buy-and-hold, everyday mutual funds (obviously Warren Buffet should be regarded as an exception...). Therefore I played around a bit with Excel and some historic data and came up with a rather simple strategy. In a nutshell the goal is to place a number of small bets on different markets that show some technical promise. Naturally there is a clear exit strategy as well, so that (the many) small losses will be cut. The goal is to ride the big waves (dotcom bubble, subprime market bubble, the recent liquidity rally etc.) a good deal of the way and therefore make up  for the small losses incurred from wrong buying signals. Back testing has shown that my strategy indeed does have a positive expactation for each trade. In theory this leaves the trade with a classic call pay-off profile, where losses are known beforehand and restricted to the premium (in my case the stopp-loss for the exit) but gains are unlimited. Another big part of trend following is conservatism (at least in terms of putting money on trades) and therefore money management is important. Without going into any further detail, my particuliar system would have generated a net return of 171% (10.49% p.a.) over the last 10 years (06.11.99-06.11.09), a time where the S&P 500 would have lost 22.5% of its value. It is noteworthy that only long trades were executed, each trade had 1% transaction cost, gains were taxed with 25% (thank you German government for that...) and investments were made in the following markets:

Dax 30, FTSE 100, CAC 40, DJA 30, NASDAQ 100, S&P 500, Nikkei 225

Hang Seng, BOVESPA, Tel Aviv 100, Mexico

Aluminum, Copper, Gold, Palladin, Uranium, Zinc

Corn, Thai Rice, Wheat, Crude Oil, Cotton

US 5 yr treasury, US 30 yr treasury

Unfortunately not all of those markets are accessible to the private investor like me at a decent cost but the investment universe is still huge and this selection still seems like a decent representation to me.

Alright, now it is time for some real life actions. I realize after last year's liquidity rally markets might be saturated and my strategy might leave me with a lot of wrong buying signals, but then again, it did work right after the dotcom bubble. So with the believe that the market is always right and price is the only indicator of its direction, I entered long positions in the following (as of 04.01.10):

Dax 30, FTSE 100, NASDAQ 100, S&P 500, DJA 30, CAC 40, Nikkei 225, DJ Euro Stoxx 50

S&P Nifty (India), MSCI Brazil, MSCI Turkey, MSCI Korea, MSCI Greece, MSCI Malaysia

Zinc, Copper, Lead, Nickel

Corn, Cotton, Sugar

With some of the trades I am rather sceptical but there should not be room for emotions in trading. That is why a technical trading system is so beautiful: Once it is running, you just act according to the rules set before hand...

With that being said, good trades and a succesful new decade to everyone!

Seb