The Hurst Trader

A very common question that I have been asked by the many people involved in beta testing THT has been “what will happen when beta testing closes?”

Well, that time has arrived, and here are the answers to those questions.

First of all, by “closing beta testing”, all that is effectively happening is that newcomers to THT will not join the beta testing group. Newcomers will be working with THT as users of the software, and not as beta testers.

I have in fact closed beta testing before – in August of 2009, when we reached 50 beta testers, and I thought that was as many as I could handle, but as the error reports started to diminish I opened the beta testing group again.

And so what is different now? The difference now is that THT is moving into a new stage of development. I am calling this new stage supported development. 

Beta testing (of new and existing features) will in fact continue, but in a slightly different way: it will be more structured. No more random “testing” of THT by simply using it. Tasks are being defined, and beta testers will be required to choose a task to work on. These tasks will not be onerous in any way, but will provide for a much more thorough testing process. Beta testers will be required to report on the progress of their testing. Many of our beta testers have been doing this anyway, and so it might not be a big change for them.

Supported Development
The Hurst Trader started as a small personal project of mine. I wanted to create software that I could use for my own trading. The project has grown tremendously over the last 9 months, since I started beta testing with the 11 people who had stumbled across this blog. There are now 300 people who have “signed up” as beta testers, and that number is growing at an accelerating pace.

The sheer volume of questions and support requests has grown to a level where I have little time left to develop new features in the software. And developing THT further is something that I am very committed to doing.

And so I have had to make some decisions about the future of THT. These are some of the approaches I considered:

Should I say thank you to everyone involved and withdraw THT from the public, and use it to trade myself? I admit this is a tempting option. It is ironic that working on THT has taken up so much of my time that I spend very little time trading nowadays! But I have decided not to do this because I am very grateful to the many people who have helped to develop THT and would like to continue this journey with them.

Should I close Beta Testing and allow all existing beta testers to use the software for free forever? I did seriously consider this option. But there are two reasons why this is not what I will be doing. Firstly because there are many people who have not yet discovered THT, and the developing community has been one of the most rewarding aspects of this project for me. I don’t want to “close the doors” on the rest of that community. The second reason is because I want to keep developing THT. I have big plans for THT, including intraday analysis, a realtime version, and incorporating a broader and more advanced cyclic analysis.

Should I “bring THT to market” and start selling it as a commercial software application? This was perhaps the most obvious option, and it is my intention to do this eventually, but it seems premature to do it now, when I have so many new features that I still want to work into the software.

And so I convinced my partners in Fortuna Software to delay “bringing THT to market”, and to rather start a period of supported development. Development will continue, and it will be supported by the contributions of THT users.

I have seen very clear evidence (emailed to me privately by many beta testers) that THT can be used to enhance trading profits. And so it is time that this project becomes self-sustaining.

The further development of THT will be supported by the purchasing of 3-month licenses. All newcomers to our user group will need to do this after the 30-day free trial expires, and existing users will transition to this new phase now.

It is my sincere hope that this transition goes well, and that a very contented group of beta testers who have been using the software for free will become a very contented group of users who know that they are getting more benefit from the software than they are contributing towards its development.

I have written before about the “grand plan” for the screensaver, and I’m pleased to announce that the final manifestation of the screensaver has now arrived.

What does the screensaver do?
Apart from the obvious: the screensaver uses your computer’s processing power when you are not using it, in order to:

• Perform backtests on a data series, according to the trading settings that were active on the chart at the time that you assigned the chart to the screensaver. This is an important point – you can test different trading settings by assigning the same chart to the screensaver several times.
• THT runs TWO backtests for every cycle of the nominal model (from 5 days to 18 years) that it finds in the data. One of the backtests uses the Nominal Model as the basis for the analysis, the other uses the Initial Cyclic Model as the basis for the analyis.

How to use the screensaver
The best way of explaining is by example, and so here is an example of using the screensaver, using the S&P 500. I am going to try to make this as easy to read as possible, so I am going to use many pictures, instead of long dense explanations!

• First of all I downloaded historic data for the S&P 500 from Yahoo Finance (it is reliable data, and free!). Simply click on the “download to spreadsheet” button at the bottom of the page, then change the name of the table.csv file to something more meaningful, such as sp500.csv.
• Then, very importantly I set the analysis start date. I have written about this several times on the Hurst Trader forum, and also spoken about it in several webinars. Basically you should set the start of analysis to just before a trough which you consider to be of some importance. I set the analysis start date to mid-July 2007, because I believe that the trough in mid-August 2007 is a trough of the 54-month cycle.
• There is something else that you should understand about the screensaver and the analysis period (how much data you assign) – it doesn’t start trading from the very beginning of the data, for fairly obvious reasons: it needs some data to perform an analysis in order to make trading decisions. THT will use anything from 80 to 400 bars to perform its initial analysis, before trading. In this example THT will use 400 bars, and so I don’t expect to see any trades before about August 2008:
  

  Note the first 400 bars will NOT be traded

• Then set all your other trading settings. In this example I am using an Expert model because that enables THT to identify longer cycles in the data than it would have otherwise. Using this expert model it will find troughs of cycles up to 54 months, whereas if I hadn’t used the Expert model, it would only have found troughs of cycles up to 20 weeks long. (I am using such a short analysis period so that I could generate this example quickly – normally one would use more data). To understand more about Expert models, take a look at the expert model webinar. The benefit of finding longer cycles is that it improves THT’s underlying trend calculations (see webinar #10)
• Then assign this chart to the screensaver:
  

  Assign to screensaver

  If your screensaver is processing some other data, and you want it to move straight on to this new data, simply click on:

• 

  Reset the screensaver to have it start a new sequence of backtests

  Note that you can assign many charts to the screensaver. It will queue them.

• Then take a break! Let your screensaver run. It will shut down every hour, so as to purge the vast amount of information that it builds up, but will start up again after the standard screensaver delay.
• Next is the fun part – while you’ve been drinking tea, or enjoying your life in other ways, THT has been working hard. On my computer this took 4 hours to complete – obviously with more data you can expect it to take much longer (see why I chose a short analysis period?)

Importing the trade histories
Now the fun begins. THT has been working as your trading research assistant, and has some interesting information about which cycles would have been best to trade over the time period that you specified as the analysis period (less the first 400 bars…)

• Open the trading history list for this data file (you can only do this from within a chart):
  

  Import a trading history

• Note that THT presents a Nominal model and Initial Cyclic model history for each cycle that it found in the data. Choose one that you’d like to look at:
  

  A summary of trading histories

  There are two side-points worth noting here:

• • The Nominal model has performed much better than the Initial Cyclic Model on this data series. This will not always be the case, and is hardly surprising in this example, because the S&P 500 is an index, hence an average, and should be close to the Nominal model.
  • The 5-day cycle is not a reliable cycle to trade! I know many people like to trade short-term, but I have found that the 5-day cycle seldom offers reliable results.
• This particular trading history includes an Open trade at the end of the history – note how THT creates exit orders, and will start monitoring this trade. If you ran the screensaver a while ago then you will probably have to exit this trade manually.
  

  Open trade with exit levels

• But why stop there? One of the questions I often get asked is “what if I were to trade more than one cycle at the same time?” Well, let’s see. I quite like the 40-week Nominal Model results as well:
  

  Summary of 40-week trading cycle

  Note that you can import many trade histories, and THT will handle them all in the same chart, but in order to trade more than one cycle you will have to create a new chart for each cycle.

• Next we will have a look at how we would have fared trading the 40-day and 40-week cycles between about August/September 2008 and November 2009.

The Trading Reports
So how would we have performed, trading the 40-day and 40-week cycles?

First of all, a disclaimer: Please bear in mind that I am presenting hypothetical results here. I am not making any representation that your trading account would have experienced the same profit or loss. Also remember that past performance of a trading method is no guarantee of future performance.

And so, let me rephrase that : Hypothetically, how would THT have performed?

• Open the Trading Report for the chart:
  

  View Trade Reports

  Note the first 400 bars will NOT be traded

• 

  The trading report

  There is a lot of information there, which I will explain in webinars, tutorials, and on this forum in the fullness of time, but for now, a few points of interest:

• • The most important number in the whole report is possibly the first one that I have highlighted – Profit Factor Two. This is the ratio between the average win and the average loss. Note that in fact THT was only right about the trades it made about half the time, but … the average winning trade (when THT was “right”) was more than 3 times as big as the average losing trade (when THT was “wrong”).
  • I have also highlighted the Annualised % return, because that is a figure that means something to me. It is certainly more than I would get by having my money sitting in a bank. Bear in mind that this return is achieved by risking no more that 2% of equity per trade (in theory – gaps and other market events could make that number higher).
  • Finally notice a very important number – the Maximum % drawdown. This is the maximum amount that the account went down (experienced a run of losses) during the period. In this example that amount is 7.77% (on the 2nd of September 2009). This is a useful figure, because you should always assume that from the moment you start trading, you will hit that maximum drawdown period straight away. Most people would recommend that you double that figure, so in this example you should assume that it is quite possible that you will lose about 16% of your equity before you make any money.
• Finally let’s take a look at the Equity graph:
  

  The equity graph

  The blue line is the actual equity for every day. The thin red line is the maximum intraday drawdown, and the thin green line is the maximum intraday gain.

• Note how the maximum % drawdown is visible on the chart. That 8% drop in equity would have required a good mental attitude (trading psychology).

Well that wraps up the use of the screensaver in THT. Two parting comments:

• Opening trading histories generated before September is not likely to work. THT will try to update the file format, but very old history files will be incompatible. In any case the phasing analysis of THT at that time was unreliable, and so the trading history files are meaningless anyway.
• Remember that trading histories built by the screensaver are the result of a purely automated process, and represent a “worst case scenario” because THT takes the most aggressive trading approach of entirely trusting its phasing analysis, and entering trades without any reference to a consideration of the market as a whole, and countless other factors that an experienced analyst and trader would be able to apply. For this reason I am actually against this sort of automated backtesting! But I also recognise that there is a place for it, so long as the results are taken with a “pinch of salt”, an understanding of the limitations of a purely automated process. A better process I believe would be stepping through the data bar-by-bar and applying some of your own experience and knowledge to the situation. (Of course you can do this with THT, but it is more time consuming).

I have realised that many people are reading this “development blog” about The Hurst Trader, instead of connecting to our Hurst Trader Forum, and so I am going to be doing some dual posting between the two, for the benefit of those who don’t go for the whole forum thing!

The Hurst Trader has an extensive “trading module” built into it (see Webinar #10 in which I introduce the way in which THT does its trading).

However I am pleased to announce a brand new feature within THT, which is a “manual trading” capability.

This allows you to:

• Tell THT that you have entered a trade.
  • you can have THT monitor the trade for you, in terms of adjusting stops, setting Take Profit orders, etc.
  • OR you can have THT leave the trade alone, in which case you will need to calculate your own exits.
• Tell THT that you have exited a trade (even though there were no exit orders triggered).
• Generate your own Action Signals (take a look at Webinar #10 if you don’t know what an Action Signal is)
  • The Action Signal can be handled by THT as a regular Action Signal that it has generated, in which case it will generate Trade Orders as necessary when the Action Signal is triggered (or before, depending on the type of Action Signal).
  • Or the Action Signal could simply trigger an alert – a message box that will pop up when the Action Signal is triggered (this might be more appropriately called a “Signal”, but the “Action” involved is triggering an alert, so I won’t get complicated about it!)
• Delete an open trade (useful if THT is monitoring an open trade, but you would rather it was finding new entry opportunities).
• Create your own Trade Orders – instead of telling THT that you have entered or exited a trade, or creating Action Signals that will have THT generate orders, you can create your own entry and exit orders, which THT will monitor in the same way as it monitors the trade orders that it generates itself.
• Add closed (historic) trades yourself, and build a trading history. If you’ve been trading an instrument for a while (before THT was available!), then you can manually enter the trading history of the instrument, and thereby use THT’s Trade Report to tell you how you’re doing (maximum draw-down, annualised % return and other useful statistics that THT calculates), including spending much rewarding feel-good time staring at the equity graph!
• Import a Trading History from a trading history file that has been built by your THT screensaver! I am very excited about this, because I have finally moved the screensaver up to its final manifestation. I will explain this in more detail in a future post.
• Clear the full trading history (so that you can look at trading various cycles, settings and so forth, and then work with a chart that represents a full history of trading in that manner).

This functionality is going to take me a few posts to explain in detail, so watch this space …

Those who are testing the ALPHA version of THT will have the functionality available in version 1.2.0.0 (available 5 December 2009).

Those who are working with the BETA version of THT can expect to have this functionality as soon as the bugs are worked out of the system (probably by about the 16th of December 2009).

I’m an apalling blogger, I admit that. Having received several encouraging emails from people interested in Hurst’s theories, I realised that I haven’t posted anything here for a while. In fact I have been working away, burning the midnight oil trying to get the software to really do what I set out to have it do: analyse the market the way Hurst would have done. I’m pleased to say I’ve made some progress. Now I am working on updating the screensaver aspect of the software so that it performs exactly as the main software works. On the way I have encountered many interesting problems, and I’m going to start sharing those on this site, which I guess is what blogging is all about! Today I’m going to write about the whole screensaver idea.

The screensaver

I have been asked why on earth I’m building a screensaver as a part of the software package. The reason is very simple. At the moment when loading data and creating a chart for the first time the software takes about 4 minutes to perform the phasing analysis on a chart with 15 years of history (on my computer, which is fairly fast, but not a super-computer by any means). Then every day it updates the analysis, a process that takes about 30 seconds. Twice in every trading cycle (and I am mostly looking at the 20 and 40 day cycles at the moment) the software does a complete re-evaluation of its phasing analysis, a 3-4 minute process. It’s all very well stepping the software through its paces every day, and seeing how it performs, but that doesn’t help one to know that in fact the software is “working” – in other words giving good advice about when to buy and when to sell. The only way to do that is to look back at a history of several years worth of the software doing its thing. To generate that history it is necessary to step the software through the data day-by-day, which is a process that really ties up your computer, and isn’t very entertaining. Which is why I am building the screensaver. From the software you allocate a chart you are working on to the screensaver, and then when you take a break the screensaver starts building the history. It’s a process that seems to work fairly well. The screensaver also builds a “trading journal” so that you can go back and see why the software bought and sold when it did. For me this is the key to making the software really work. Of course I can only expect the software to work as well as an expert in Hurst’s theories would work … and although there is some evidence that the theory works, the real test will be the trading history generated by the software, and the ongoing trading the software does.

Backtesting

Backtesting has earned itself a very bad reputation in trading circles, mostly because it is possible to optimize the parameters a system uses so as to perfectly fit the historical data, and produce a fabulous trading history that simply doesn’t work when you use the same parameters on new data. So is there a danger that my screensaver is going to fall prey to this dreaded problem (which has been the downfall of many trading systems – and is related to the problems Rich Swannell encountered with his Refined Elliott Trader because he presented hypothetical results rather than real ones)?

The simple answer is no! But I should explain that. There are two reasons why I don’t think optimized backtesting is going to distort the results generated by the screensaver:

•  There are no parameters involved in what my software is doing. I am not optimizing any parameter values. The software performs phasing analysis in a way very similar to the way Hurst prescribed – by using a visual “pattern recognition” approach rather than a mathematical analysis, and so there are no parameters to optimize. The software won’t be changing what it does (or the way it does it) because of the data it encounters. I will write soon about how the software performs phasing analysis because that is key to what it does.
• Secondly the screen saver actually works with truncated data – even though the data for the “future” is available to it when backtesting, it truncates that data and works only with data up to the day it is testing. This means that any mathematical calculations (such as moving average calculations) cannot use data that would not be available if you were running the software on the date in question.

There are two ways in which the screensaver’s results will not match the real performance of the software:

• I am not using “tick” data, and so the software doesn’t detect situations where a market moves too fast for you to get in or out at a particular price. If the market traded at a certain price (it is included between the high and low price of the day) then the software assumes you traded at that price. Of course when actually trading there is often “slippage” – failure to trade exactly at a particular price.
• The other difference is that there is no human involved in the testing process. As much as I have great respect for humans, they can make some really foolish trading decisions – and when back testing the software doesn’t have a human filter saying “Buy? What nonsense … there’s obviously something wrong with this stupid software. I’m certainly not buying now.” Right before the biggest rise in the history of the stock you’re trading. I know it happens – I’m just such a human!

OK, so that’s it – why I’m creating a screensaver. The other thing about it is I think it looks cool! Here is a sample:

 

screensaver in action

I keep writing all sorts of nonsense on this blog, but the one thing I haven’t got to yet is what the software is doing right now. That’s because I’ve thought I needed to prepare the ground, explain things, describe the trading theory at length … but instead I’m just going to take the plunge. Pictures say it all:

This is a chart of the Euro-USD exchange rate, produced by the software, analysing data up to Friday 30th May 2008.

• The thick BLUE line is the “price” line – the exchange rate between the two currencies.
• To the right you will see the price scale, with the closing price of Friday 30th May highlighted in a red box – it is red because the price has FALLEN (is lower) from the previous day’s closing price.
• The thin multi-coloured lines are the lines Hurst calls FLD’s – Future Lines of Demarcation.
• Price is surrounded by a “dominant channel”  the purple dotted lines that form an envelope around the price movement.
• Along the bottom of the chart is the last part of the phasing analysis for this chart. To the right you will see the cluster of small circles within lines, with a grey shaded background. This indicates that we expect a nest-of-lows within the next 7-10 days, with a strength of about 20 weeks.
• You will also see some VTL’s – Valid Trend Lines, as defined by Hurst. These are the straight dotted lines, with small boxed labels giving the wave period of the VTL.
• Notice also the TRIO of moving averages, moving around the price line. One is dark red, one blue and one green. These lines are not Hurst-inspired, but are a personal refinement by which I calculate refined FLD projections.

I’m not going to explain at length what this chart means because that would take several pages of discussion. For now I just wanted to give a glimpse of what the software is doing. Having said that, here is the commentary generated by the software, which pretty much does that for me:

Updated Phasing Analysis reveals that the dominant cycle has a wavelength of 16.9w, or 118 days. This is assumed to be the 20w nominal wave, which implies that all waves will be (-14%) shorter than nominal. This cycle is approaching a TROUGH (and a medium degree nest-of-lows). The underlying trend to this cycle is currently FLAT, likely turning DOWN. The future FLD indicates that this cycle will influence prices to peak in a short while, and then to fall.

The sub-dominant cycle (next wave down from the dominant cycle) has a wavelength of 58.9d, or 59 days. This is assumed to be the 80d nominal wave. This cycle is overdue for a PEAK. This is probably the second of 2 sub-waves. The underlying trend to this cycle is currently FLAT, likely turning UP. The future FLD indicates that this cycle will influence prices to peak soon, and then fall.

The pre-dominant cycle (next wave up from the dominant cycle) has a wavelength of 33.9w, or 237 days. This is assumed to be the 40w nominal wave. This cycle is approaching a PEAK. The underlying trend to this cycle is currently strongly UP. The future FLD indicates that this cycle will influence prices very little because despite near-term volatility the FLD becomes range-bound.

Σ L is currently FLAT. (value of: 0.0)

The last phasing analysis was 4 days ago. In that time price has FALLEN by 244 to 15521.
Sequence #1 of the previous pattern: Price did move DOWN (for 3 days) as expected and has moved towards the target of 14954. The closest price has gotten to this target so far is 15394 (2.9% short) 1 day ago. This target is now 38 days away.
This brings the price movement up to date. From this point on:
Sequence #1 is expected to reach a LOWER target of 14808(-146 diff) 4 days EARLIER, on 2008/06/12 Recent price action has generated a new sequence (# 1). The target of this projection is 13624, expected by 2008/10/06.

CURRENT PROJECTION PATTERN:
Price is currently range-bound (last close 15521), in an active FLD pattern cascading down. Price crossed below the 15 day FLD 1 bar ago, implying a price fall to 15467. This move would imply a price cross below the 29 day FLD (est: 2008/06/03 & level: 15498) projecting price to a CONGESTION PAUSE ZONE at approximately level 15436  Then price is likely to CONTINUE DOWN after a pause, towards fulfilling the 16 week cycle projection of 15402  taking price into a CONGESTION PAUSE ZONE at 15431 by about 2008/06/09  Then price is likely to CONTINUE DOWN after a pause, towards fulfilling the 59 day cycle projection of 15342  taking price into a CONGESTION PAUSE ZONE at 15429 by about 2008/06/09 . This price move breaks below the 33 week VTL implying that the peak on 2008/04/23 is the expected peak of the 15 month cycle. Then price is likely to CONTINUE DOWN after a pause, towards 14866 by 2008/08/09, attempting to fulfill the projection generated by the crossing of the 58 day FLD taking price into a CONGESTION PAUSE ZONE at 15257 by about 2008/07/03 . This price move breaks below the 16 week VTL implying that the peak on 2008/04/23 is the expected peak of the 33 week cycle. Then price is likely to CONTINUE DOWN after a pause, towards 14522 by 2008/11/09, attempting to fulfill the projection generated by the crossing of the 33 week FLD taking price into a CONGESTION PAUSE ZONE at 14782 by about 2008/10/05 . This price move breaks below the 15 month VTL implying that the peak on 2008/04/23 is the expected peak of the 46 month cycle. Then price is likely to CONTINUE DOWN after a pause, towards 13572 by 2009/07/15, attempting to fulfill the projection generated by the crossing of the 15 month FLD taking price into a FINAL REVERSAL ZONE

There is no cycle which provides an average potential of at least 20% per half-cycle, and so it is not recommended that you trade this stock unless you are trading on a leveraged basis, in which case you should change the Trading Settings by using the Tools|Trade Settings|Edit menu.

Having reviewed the entire projection sequence I am unable to find a move suitable for trading, and so no trades have been set up.

 Well – that’s what the software is “thinking” about the Euro-USD as of the 30th May. It seems pretty likely that things are going down, and that the upcoming nest-of-lows will generate little more than a temporary upward correction to the downward move. However it is early days for the software, and so I am by no means making any kind of call on this, I merely put it up here to show what the software’s doing at the moment.

I’m not a blogger – let me just put that out there right up front. I have always rather scorned the concept of making a diary “public”, but I have had an epiphany of sorts and have realized (I guess about 5 years after everyone else) what the value of a blog is, and what I hope to achieve from this blog – I want to share my ideas, and find someone or perhaps some many to talk to about them. As my wife’s eyes glaze over whenever I mention my software project, I reckon it’s about time for me to find some other(s) to talk to about it.

The Project
In August of 2005 (almost 3 years ago as I write this) I started working on a personal project which has become the subject of this blog – I decided to create software for use in trading (of stocks/ shares/ forex, you name it) which would be able to replicate all (or most) of the thought processes that go into an approach to trading that intrigued me (I’ll discuss this trading approach in detail in a later blog). I know that’s a tall order, or at least I have realized over the past 3 years what a tall order it is, but I believe I am beginning to realize the dream, and I feel the need to share the project with others who might be interested.

Because here’s the thing: I have tried a good many of the trading software packages out there, and I have been disappointed with most of them. The “black box” software has never appealed to me (so to be fair I’ve never given it a good run) because I’m the kind of person who likes to make my own decisions, particularly when my own money is on the line. And other analytical packages I have found frustrating because either:

• They are purely “charting packages” some of which allow you to design your own indicators. These are great for what they are: they are like the word-processors of the trading world – they deliver fantastic looking charts, but their greatness rests in the hands of the user! I want to create something that is beyond a tool.
• Or, they are software packages that go one step further and analyse the data according to some or other theory, but then take no “responsibility” for that analysis, by which I mean that there is no continuity from one day to the next. One day the software will be predicting a bull move, and the next day (after the market crashes heavily) the software happily tells you that it was predicting a bear market all along. How often have I agonized over a trade, finally taken the plunge, pulled the trigger, bought the stock, only to lose money and the accursed software has no memory of the false prediction it gave.

I’m a little obsessive. I have tested many software packages by stepping them day by day through the data – and I have yet to find one that really works consistently across all market conditions and several markets. And yet in my (fairly extensive) reading on the subject I have come across at least one theory about the movement of stock market/currency prices that really seems to make sense, and which should work. And hence the project that is the subject of this blog – to create software which will be able to:

• Analyse the stock market according to “the theory” – I will be writing much about the trading theory the software uses later.
• Make “intelligent” trading decisions, in terms of:
  • Trade entry (buy and sell signals)
  • Trade monitoring (including the facility to “know” why the trade was initiated in the first place)
  • Trade exit – selling (or covering short) at the appropriate time and price.
• Furthermore, the software should have some concept of continuity over time (in other words if an analysis changes it should recognise that fact!)