More Help Available

The Thinking Stuff Website has been fleshed out a little. It's not finished - I'm still going through and moving content from this blog to the website, and when you read it you'll see what I mean. Anyway, if you have, or are about to download the software, I recommend going through the pages listed under the Help menu.

Topics there are:

• General Window Flow


• Getting Started


• Setting Up To Download Price Data (overlaps a bit with Getting Started).


• How Does Thinking Stuff Decrease Risk?


• How Does Thinking Stuff Do Auto-Trading?


• How Does Verification Work?


• My Chart Window Doesn't Open


• Available Trading Rules


• How To Open A Gain Capital Account (nothing new if you already read the blog entry).


• How To Install PostgreSQL (also nothing new).


• Known Bugs

The most notable absent Help topic is how to set up trading systems. This topic is touched upon in the "How Does Thinking Stuff Do Auto-Trading" page, but I want to make a page which has a lot of different examples. It's coming soon.

Reply To Luca

Yay! Received a comment on my software:

Greetings Mark,
thank you for sharing the knowledge and allowing free testing of your software. I haven't got round to actual backtesting or autotrading, couldn't download the data from GAIN. I currently use MT4, lot more complicated since it needs real code, but quite easy once you get the hang of it.

I find your software very easy to install/setup and to desing a system. Of course the fact that it is user-friendly has limitations, but very cool for non-programmers. I feel the money management and position sizing are a bit limited, and backtesting on only one contract sucks. And it is limited in the way you can change the exit rules depending on the way the trade "behaves". I haven't tried, but it seems you can backtest only on one currency at a time. Not good! And of course the analysis of the backtest results are super important. If you want to give yourself a key differentiator in the market, based on my limited experience, correct portfolio backtesting and system performance analysis reports would be a good area to spend time on.

Thanks again,
Regards,
Luca from France

I really am appreciative that someone took the time to send me such a detailed comment. But there are some parts of it that I feel I have to clarify. It'll sound like I'm being defensive, which is true :-)

Firstly, the comment "I find your software very easy to install/setup and to desing a system. Of course the fact that it is user-friendly has limitations, but very cool for non-programmers" is very, very true. It does have limitations in that you can only select rules out of those I have coded for you. However, if you do want a different rule coded for you, just send me an email with the details.

"I feel the money management and position sizing are a bit limited..." The software has the following money management options:

Last Gasp settings:

• Pass on any trade where the price goes X pips past the calculated entry point.


• Pass on any trade when we can't get in at exactly the calculated entry price.


• Anyting is okay.

Number of Pips At Risk settings:

• Pass on any trade where the risk is more than X pips.


• Anyting is okay.

And with auto-trading there are the following Position Sizing options:

• Adjust number of units purchased so that less than X% of balance is at risk.


• Always purchase X units (100,000 units equals 1 contract). And then if you choose this option you also choose one of the following % Of Bank At Risk settings:
  
  • Pass on any trade where more than X% of balance is at risk.
  
  
  • Anything is okay.

I'm not sure what other options are possible. Either you have a position size based on a % of your balance, or you have a fixed position size. If you can think of something else I'll code it.

"backtesting on only one contract sucks..." This was done purposely so you are always comparing apples with apples. I think I found this tip on the Currency Secrets website, but after a quick look I couldn't find the exact webpage.

Backtesting is done so you can compare different trading systems against each other. Or, to compare the same trading system with slightly different variables. If you backtest with variable contract sizes then it won't be a true comparison.

"And it is limited in the way you can change the exit rules depending on the way the trade "behaves"... I don't really understand this statement. Could you expand a bit Luca?

"I haven't tried, but it seems you can backtest only on one currency at a time. Not good!" Actually this statement is very, very *not* true. You can trade as many currencies as you like in the one backtesting session (that is, the one click of the "Start" button on the backtesting window). You can even backtest over different time periods in the one session. And you can backtest many different trading systems in the one session.

"...correct portfolio backtesting and system performance analysis reports would be a good area to spend time on..." Just let me know what you want and I'll code it.

Right now the backtesting results show:

• # Trades


• Pips Profit


• Profit Factor


• Start USD


• End USD


• % Gain


• % Gain p.a.


• # Wins


• # Losses


• # Break Even


• Pips Won


• Pips Lost


• Pips Risked


• Risk To Reward


• Av Pips Profit


• Av Pips Won


• Av Pips Lost


• Av Pips Risked


• % Wins


• % Losses


• % Break Even


• Max Winners In A Row


• Max Losers In A Row


• Slippage Included


• Pessimistic Trades


• MAE


• MFE


• Max Drawdown


• % Drawdown


• Recovery Factor

All of the above are shown for all trades, for short trades, and for long trades. Along with that summary, the individual trade details are also listed.

I hope that's put your mind at ease. Thanks again for the comment Luca. Keep 'em coming!

Back

Alrighty, back home now.

I see that about 20 people have downloaded the beta-version of the software. What do you think? Any comments? Good or bad things to tell me, it's all good.

Just leave a comment on this page, or use the Contact page to send me an email.

Okay, Go Get It!

Well, I didn't quite get to finish as much of the webpages as possible. In particular, most of the pages under the Help menu are not done (on one page you should even be able to tell the exact point that I got tired and decided to go to bed).

But, the main page is there, and you can download the software from that. Installation requirements and instructions are also there. Go to www.thinkingstuff.com.

Unfortunately, my girlfriend is coming to Australia for a 3 week holiday, and tomorrow I have to drive 500km to go and pick her up. Then I have to spend 5 days at the beach with her.

It's a tough life.

But, I should be able to check my email once a day. If you have any questions that aren't answered on the new webpages (highly likely), just shoot me an email.

Nearly There...

Well, a day or two and I'll release the Thinking Stuff software, in its beta form, to anyone who wants it.

Being beta software, I hope you realise that it will have bugs in it, and there will be new versions released every so often.

Just let me create some rudimentary webpages so you can download it, and know how to install it and get started.

It's One Week Later

From previous post: "So... a week later... if there's nothing majorly wrong with it (that we know about)... I'll make the pre-release version available to those who want to start using it as soon as possible."

I gave the first version, v0.6.0, to my beta testers about a week ago. A couple of days after that I sent v0.6.1, and the day after that v0.6.2.

The problems encountered were mostly along the lines of getting their Thinking Stuff database set up. That is, nothing actually to do with the currency trading part. Because they couldn't get that far. It was related to the fact that I had switched to PostgreSQL at such a late stage.

If you read this webpage and scroll to the bottom, you can see what I fixed in the 2 newer versions.

Now they can get past the Login page, the beta testers are generally a little confused as to what to do next. I purposely gave them no instruction to see if they could work it out. But they had troubles.

Ironically, I thought my GUI design approach, while not "standard", would be easier for people to use. I know that keeping to standard industry approaches are preferred, but it seems to me that the standard approach is not the best one. And, because people keep the standard, in effect we are keeping ourselves, on purpose, attached to something sub-standard. Just because "that's what we've always done". I hate those words.

So instead of one-word menu items, like "File", "Edit", etc, I tried to change the world a bit with "I Want To...".

And then the menu items leading off that were "View/Change...", "Manually...", "Start/Stop...", etc.

And the menu items leading off "View/Change..." were "View Charts", "View/Change accounts", "View/Change what data is downloaded and what is created", etc.

With each sub menu item it builds an actual English sentence.

I think that part might be fine, but I also have "I Want To..." menus on the bottom left of each individual window. On those you see items like "Add a new item", "Save and close the window", "Save and keep the window open", etc. My beta-testers were very attached to their "Save" buttons, and the Xs at the top right hand corner of the windows, because that's what all other programs have.

I think I'll put the Xs back (except on one window that I don't people to be able to close). But, in general, whenever you want to do something, just look for something that says "I Want To...".

I will also create some kind of wizard that will guide people through setting themselves up in order to connect to Gain Capital. And create a webpage to help people get started, which I planned to do anyway.

So, although I know there are people who are ready and waiting to go, I hope they can forgive me that I want to wait until the beta testers have covered the currency trading aspect of the program before I make it available to the general public. That will also give me time to create some basic webpages with instructions.

But, I have to say sorry if I misinformed anyone over the cost. When the auto-trading functionality has its bugs mostly ironed out, it will be US$30 per Month.

Until that time however, it will be free. Free to anyone who signs up for the backtesting functionality, which is US$200 per Year.

That way I can get more people using the auto-trading, and therefore hopefully any bugs will be pointed out more quickly.

Of course I'll give lots of notice to people who take me up on this offer, as to when the cost of $30/Month will start.

And of course, if you really really want to get started right now, or you really really just want to use the auto-trading, and you really really don't mind using a beta (buggy) product, nor the installation of a new version every couple of days, then please just contact me and we'll get you started. I guess I should set up some way for people to pay me...

Summary: 1 more week people! But the product gets better and better, thanks to my lovely beta testers.