- Overall stability
- Performance and capacity to handle more volume and operations
- Better graphics
To take 4.0 for free trial click here
Some of the new stuff in 4.0:
- Database change to Microsoft SQL Express: more data & increased overall performance.
- New graphics: faster rendering, mouse/trackball control, dynamic data labels and more robust
- New back tester: more info, new P&L chart presentation, better out-of-sample testing
- New Portfolio back tester: support for intraday portfolios, new charts and and benchmark performance
- New user balancing mechanism: user choice to override auto-hedge
- New ROIC calculation: implemented in both individual and the Portfolio back testers
- One click enabling of Scheduler jobs
- Faster start-up initialization
- Updated screenshots here
- Updated documentation here
Conventional wisdom suggests that weak out-of-sample results mean unreliable in-sample results. Yet that is not necessarily the case: in-sample data is not always biased to detect spurious predictability.
For example, splitting a sample between in and out portions may mean a general loss of information and less predictive power. This is aggravated in samples of small size. Thus an out-of-sample test may end up falsely rejecting valid in-sample results.
There is also the issue, in the context of our software, of reconciling instances where the split period used for the training turns out not to be a period over which the pair is cointegrated whereas the whole period is.
Overall the message might be this: traders could usefully take the view that both in-sample and out-of-sample test results be viewed as informing the pursuit of tradable pairs rather than one being the underwriter of the other.
Check out the screen shots from upcoming version 4.0 here (images 20 to 25) to see how ArbMaker performs out-of-sample tests: .
A common question for us is ‘why do betas in ArbMaker differ from those on financial websites?’
The typical beta calculation on, say Google Finance, uses the S&P500 as a reference point and is calculated via a regression of the returns data of both time series. This approach is at the heart of correlation.
The regressions in the Engle-Granger model of cointegration inside ArbMaker are calculated using price data – not returns. That emphasis points to a key difference between correaltion and cointegration; and means the resulting betas cannot be compared on a like-for-like basis.
The focus under cointegration on price data has distinct advantages over the returns-only approach for spread analysis. That is because there is comparatively limited information in returns data whereas cointegration extracts information from price data at several levels. Price data, for example, is highly autocorrelated and yields precious information about spread trends, direction and persistence that is lost to returns-based data alone.
Yet both betas have relevance – it is the key ratio defining the amount per side when configuring a pairs trade. In our next release, due out in July, we allow users to tweak this ratio. All the more reason to grasp the important differences between the two betas!
A few things to expect:
- New graphics upgrade supports fast dynamic updating at all intervals (intra and end-of-day). Couple of examples here (back-tester) and here (Bollinger and Indexed Spread).
- Don’t like the hedge ratio and want to tweak it? Go ahead with the manual balancing function
- Support for intraday portfolio tests alongside the existing end-of-day portfolio test
- Moving to SQL Express database for greater stability
- More controls in the dynamic filter section of the Tracker for finer statistical profile requirements
- Return On Capital Invested ratio at the trade and portfolio level
- Improved internal process flow for speed
- Introduction of a ‘FX plus Scheduler’ version
- Assorted bug fixes including the pop-ups behind the main window and full Windows 8.1 compatibility
If you are already a client and want to beta test this version NOW please contact us.
This FX performance using our software was sent in by the Head of FX & Metals for Gleneagle Securities and Global Prime in Australia. Impressive, especially on the drawdown and PIP count.
Gleneagle intend to run the model in this live account another quarter and then assess their next steps – possibly including closer collaboration with us and a Multi Account Manager offering.
In our latest newsletter we include a case study of a real trade recorded as it unfolded from start to finish. The study sets out and explains the settings/parameters used and provides narrative for the context of the transaction itself. It is useful reference information for newcomers to the software; and may provoke a few ideas for long standing clients too.
Click through on the graphic below or here to get to the download link:
Like we say in client calls, the stat merely tells you something might be worth checking out. Even deterministic arbs like this one (ie “can’t miss”) as well as relative value opportunities appreciate some due diligence.
Our latest version for non-FX versions is now released. If you are a client and have not received your update email you should get in contact with us.
The main changes:
- major reworking of the feed connection code in the Tracker to handle more volume and refresh faster under that volume
- bid/ask ratio implementation for the latest price in the Tracker gives more accurate signals
- additional Scheduler information to cover number of trades and biggest wining trade
- a revised drawdown metric in the back tester now calculates on an intra-trade basis rather than only the entry and exit values
- new trade duration metric in the back tester
- expanded export options now include configuration parameters (brokers, feeds, APIs, exchanges etc)
- assorted bug fixes and refinements
We have a long wish-list of requests we are also working on for a late-January release. What we could not include this time (for example, the dollar value per standard deviation, scaled entries, eSignal connection etc) we continue to get ready for future release.
If you are a demo user consider booking an appointment with us to check out and gain access to this latest full unrestricted version. It is light years ahead in terms of speed and functionality.
Lastly, it is that time of year again: during the Team’s holiday break 13 December to 6 January we will provide emergency support only.
Season’s Greetings and thank you for you custom!
The October newsletter is out in sync with version 3.4. We strongly encourage interested traders to click here to book a Webinar Overview with us and after the presentation take up the offer to try it for free for 30 days.
(Just click the image below to see the newsletter).
Stimulating article a few days ago from Mr. Basenese at Wall Street Daily in which he links income inequality to a relative value pairs trade opportunity between Wal-Mart and LVMH Moet Hennessy Louis Vuitton SA. He presents this chart in making his case:
Now there is no statistical suggestion as to cointegration support for the pair over the period 2012 until now in Mr. Basenese’s analysis. Does any exist that would buttress the argument presented? We ran WMT against MC (the primary Paris listing of LVMH) to see.
There is indeed cointegration at the 90% confidence level since 2011 – but not when using 2012 as the starting point. That may suggest the relationship is weakening. However, whilst trading a pair like this runs legging risk – the primary exchanges are not open concurrently – the P&L record over both 2011 and 2012 is still good:
Visually the trades panned out this way:
Some details of the analysis:
- End-of-day data
- Basic Z-score based strategy under a fixed beta calculation method
- 41 day average holding period per trade
- Outlay per trade = ~$9000 unmargined, ~$3600 margined
- Gross profit per trade $565, a return per trade of ~6% unmargined and ~15% margined
- Margin assumed at 40%
- Unmargined profit = 51% over the period
- Margined profit = 128% over the period
- Currency differences accounted for
- Trading hour differences accounted for
- Relative beta accounted for
The data suggests taking a long position today in Walmart and shorting LVMH. But, given the deteriorating cointegration profile 2011 vs 2012, how much longer will the theory and the stat continue to hold and support the trade…