Starting a new series, using the demo-generator

After a long silent period it’s time to start another series of blog-posts. During the summer most potential  readers were at the beach, sailing or traveling through a foreign country, not thinking about business and work at all. So why bother?

Of course I discussed a lot of different annual reports, comparing  different companies or a  company through a series. Often it was a combination: several companies and a couple of years.  E.g: about banks in the Netherlands, about oil & gas companies,  about very different industries in comparison,  again about  very different industries in comparison but this time really large ones, usually  well-known (or at least their products and services).

For all these examples a full licence was needed, but the demo-version of AnRep3D is free. Why not use our own free demo-package  for a change, to  generate graphs the way everybody could do with the free demo-generator? Well, next time I will present a couple of companies (only one company a time for just one specific year as the free demo will generate only one “building” in the graph). Don’t think a graph generator with the free demo-generator wouldn’t be interesting!

Because four values are combined in one single building (total assets, equity, revenue and profit) a lot of ratio’s can be deducted visually: revenue or profit compared to equity, revenue and profit, equity to total assets. All relative measure providing interesting insight.

Let’s try and work out a couple of companies next time. Hope you will like it and don’t forget: the demo-package can be downloaded for free. The generator comes with the manual and examples of input- and output-files (graphs, in html format, so basically webpages your browser can read and show). For now it’s available without any registration of name, email-address or phone number! No cookies either – completely anonymous, because we want everybody to have a look:

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , | Leave a comment

Why we had to drop the stereo-view for AnRep3D

After all the examples given, showing the usage of the AnRep3D generator, it’s time for some retrospective. It will be bit of a technical discussion, so if you don’t like it, just skip this post as the subject is a one-off for now.

The original plan was – and still is – to create a Virtual Reality environment, where the graphs can be walked through and manipulated in real time. Actually, our first step was closer to this objective than the current implementation of AnRep3D. The old versions of the generator were providing VRML output. VRML stands for Virtual Reality Modelling Language (in the past the formal expression was Virtual Reality Markup Language). It became popular with a small group of people and organisations only and as a result there was e.g. no native support for the language in browsers or other broadly used applications. Plug-ins and stand-alone viewers were available – even for free – but we noticed people didn’t like to go there.

VRML is a rather old language and it has a successor in XML-format: X3D. Unfortunately this new official standard is supported even less than VRML itself! So about two years ago we decided to switch to a new concept, lowering the threshold for users. Since then generator provided HTML5 code, to be read by all modern browsers. The choice was still between generating a lot of Javascript-code (I hate the language but it’s everywhere) or some kind of rather clean X3D, embedded in a HTML-page as supported by X3DOM. X3DOM is a standard and a supporting organisation (see: ). Deep down both solutions use WebGL ( ), so basically there’s no difference but X3DOM takes care of all the Javascript nuisance.

So after years of research and development using VRML, we moved to X3DOM and now the output of the generator can be presented by a web-browser without any additional actions like installing plug-ins. One of the drawbacks however, is the lack of support for stereoscopy. For VRML a very nice viewer was available (see: ) offering stereoscopic view based on a left-eye and right-eye presentation (and other options as well). For X3DOM or at least HTML5 with X3D (using WebGL) it’s not impossible, but very, very hard as the whole image has to be generated for both left and right eye – both presenting a slightly different viewpoint of course. But to be honest: the parameters for the stereoscopic viewer had to be optimised to get the best result and only experienced users would go down that road.

Yet it’s nice to see the stereoscopic version of the AnRep3D-graph using some VR-gear like wither an Oculus or a mobile phone in some cardboard-equivalent. In the picture an example of a cheap one I purchased recently.


Don’t think it’s impossible to offer stereoscopic pictures for an AnRep3D graph. Screenshots can still be stereoscopic, because the left-eye picture and the right-eye picture can be taken as separate screenshots and placed together. Remember: It will do to present them on the screen of a  mobile phone which is placed in some cheap VR-device (about 10 euro). The same way as stereoscopic Youtube movies are watched. An example of (the box of) such a device is shown to the left.


The actual screenshot-pair taken from a graph discussed in a previous post is shown below (please download the original – it’s larger and therefore the resolution will be better).

Stereo-view of graph large companies

For people who prefer to squint when looking at stereoscopic photographs (I do it myself), there is a second version of the picture with the left-eye and right-eye image switched. It has a tiny white line in the middle (on purpose, to avoid confusion with the headset version).

Crossed stereo-view for graph showing large companies


The real difficulty is having a stereoscopic image to be manipulated (translate, rotate, tilt) in real-time. Eventually we will get there again. For now it’s important everybody in the world knowing about AnRep3D and the people needing it all having a licence!

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , | Leave a comment

Visualising a mix of large companies in a spatial graph (III)

Last time we checked the values again and corrected the input for PSA. E.on however, has really shrunk! The 2016 loss was extremely high and as a result the equity almost vaporised. The total assets and the revenue are much, much lower as well. This is not about taking an annual report from the wrong company. E.on decided to transform itself, because of the German “Energiewende”. Some background information can be found in this article: What we observe in the graph is the truth! Below a screenshot clearly illustrating the change (E.on is in the middle, with 2006 in the front and 2016 to the rear).

Eon shrank

Although E.on is a striking one, the other companies are important as well, so let’s discuss them from the left to the right:

Tesco sideview

Tesco did slightly better in 2016 than in 2006. The revenue became higher and the ratio of the equity to total assets improved. The former fact can be observed in the sideview.


Top-view large mix

The latter is visible from the top. In the same screenshots we can see PSA (second from the left) improved its equity to total asset ratio much more. The large loss E.on suffered in 2016 cannot be seen from the top, but is clearly visible in the side-view above.





Of course all these screenshots are poor substitutes for the real graph, so have a look at the original 3D version. Clicking on the screenshot will do because of the hyperlink. The 3D-graph is just an HTML-page and will appear in your browser. As stated before it can be manipulated easily:

Clicking the right mouse-button, moving the mouse up and down will zoom the graph in and out. Clicking left and moving the mouse will tilt the graph in different directions (or move the observer’s viewpoint around a fixed graph – it’s relative of course). Double clicking in the graph translates it and moves the centre at the same time. As a result the way the graph tilts will change. Just try it. If you don’t know how to get the normal position back, just refresh the graph.

IBM and Maersk front-viewTo the right we still have Maersk and IBM. For both the revenue decreased a little bit between 2006 and 2016, but Maersk went from profit to loss (no wonder for logistics), whereas IBM had a healthy profit in both years!

Again, it’s not the perspective, because we see the same pattern when turning around 180 degrees. Now IBM is to the left of course and Maersk to the right.

IBM Maersk 180 degrees rotated


No need for additional comments. You can have a look for yourself and draw your own conclusions!



If you want to try it yourself just download the demo-package. It’s free – even no registration needed at the moment. Only one “building” will be shown. If you want to visualise a series of companies over a range of years (e.g. ten companies over ten years is a possibility), you will need the fully functional generator. Contact us ( or or have a look at our AnRep3D website first.

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , | Leave a comment

Visualising a mix of large companies in a spatial graph (II)

Picking up the subject where I left it last time, I have to convert all the numbers into the same currency. Otherwise we cannot compare the numbers in a spatial graph (nor in any other graph for that matter).  This time, it will be USD, so we have to convert EUR and GBP to USD. To be able to do so, I used a very useful site! As revenue and profit will cover for the whole year, I had to derive an average exchange rate, whereas equity and assets are a snapshot at the end of the year, asking for an exchange rate at the end of the year as well. I applied:

GBP  -> USD 2006 average $1,87 and for the end of year value $1,95

GBP  -> USD 2016 average $1,35 and for the end of year value $1,23

EUR -> USD 2006 average $1,27 and for the end of year value $1,32

EUR -> USD 2016 average $1,10 and for the end of year value $1,05

One could argue the values chosen but this this blog is not science, it’s about showing the power of spatial graphs in financial analysis. So let’s move on.

Some values were a little bit odd (or even very odd, like for PSA and E.on) and therefore I double-checked the values. I found some typos, but the values which struck me as odd were alright. This means some strange things happened in the period between 2006 and 2016. One special event is known as the “financial crisis”, but other influences could apply as well, like divestments or mergers and acquisitions. My input-file looked like this:

input-file large_mix

Despite the double check: having a first look at the 3D-graph (clicking the screenshot below – or just looking at it), it really didn’t make sense!

large_mix topfront view

To the left, both E.on and PSA, large companies in 2006, look very small in 2016. If we tilt the graph a little bit more, it’s even clearer: see below. large_mix tilted

It’s not some strange parallax or perspective. If we turn the graph 180 degrees it’s not very different (below).

large_mix 180 degree trun

To understand what went on, I checked wikipedia and annual reports for other years. I will talk about E.on later,  but despite the crisis PSA (the Peugeot Citroën company) didn’t change that much. Looking for 2015 I got an annual report in South African Rand and I realised this was about another company. Yet, my numbers for 2016 were in euros and I didn’t expect another PSA to be active in Europe. On the other hand for a completely different industry it would be possible and since I only went for the income-statement and balance-sheet, I could have missed something. And I did! Asking for annual reports 2016 again, I noticed a “PSA International” just above the “Groupe PSA”. The former was a smaller fish-related company, but the latter was the one I looked for and it was all about cars indeed.

For E.on the 2015 numbers looked quite normal as well (I even checked a whole range of years), but then suddenly the 2015 numbers in the 2016 report were quite different from the original ones. After large divestments E.on became quite a different company in 2016. After correcting the PSA numbers the graph looks better. Next time I will talk about E.on and the other companies in more detail, using the new graph below (just click on the screenshot to have a look). For all companies we see the impact of the crisis, but the graph makes more sense now.

(To manipulate real 3D graph: zoom = click right button and move mouse, rotate = click left, hold button and move mouse, translate = double-click in graph. Restore the original position by using the browser’s refresh-button. )

large_mix corrected

For more information about our generator for spatial (3D) graphs of annual reports, have a look at AnRep3D

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , , , , | Leave a comment

Visualising a mix of large companies in a spatial graph (I)

Before the “abuse” of  the AnRep3D generator to visualise the average, individual energy-usage in different countries, normal AnRep3D spatial graphs were shown and discussed. With “normal” I mean visualising key financial data from the annual report (or more precisely: the Financial Statements in the Annual Report).  That’s what AnRep3D was developed for, although derivatives for other purposes could exist in the future, hence the demonstration of the “abuse”.

This time we go back to the original purpose, like we did in the series about relatively small companies. The difference is, the companies picked now are much larger. Tesla, EasyJet and USG People have revenues with a magnitude of billions, but the companies I selected this time, have annual revenues of tens of billions. At the same time, the AnRep3D-graphs are relative and a graph comparing small companies only won’t be very different from a graph comparing large companies. In a mix the difference would be huge, of course. Much like we saw with the energy-graphs, where the USA and India were in the same graph.

AnRep3D is all about comparing companies by looking at relative dimensions. It’s not important how large a company is, but the ratio of profit to revenue is, as is the ratio of equity and total assets (and e.g. the ratio of the revenue to the total assets for that matter). Even more important is the possibility to see companies change over time.

For this new series, I picked five very large companies in Europe and the USA: Tesco, PSA, E.on, Maersk and IBM. The most recent year – 2016 – is interesting but 2006, before the onset of the crisis, would make a good comparison. So I had to select ten sets of data: two years for five companies. Each set consists of four values: revenue, profit, equity and (total) assets. The abbreviation being R-PEA (i don’t know a real “pea” of this type, but its rather easy to remember).

For now I won’t present anything, except for the raw data I took from the annual reports (still a mix of USD, GBP and EUR). Some annual reports have a slightly unusual format and I don’t hope I made a mistake in picking the right values. Yet a warning: don’t take my numbers to work with when looking for financial information. Always go back to the source and look for yourself. If you want to visualise the data, please purchase an AnRep3D  licence 🙂 or download the free demo-set (only one “building”) at

raw input data large companies

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , , , | Leave a comment

Abusing the AnRep3D-generator: visualising energy-sources (IV)

The last time I didn’t provide the readers with a real spatial graph, because it had not been uploaded yet. During the week after the post, I realised the full power of the AnRep3D generator – even if it’s abused for energy instead of money – needs more than just a single year. We already had eight countries for which we showed the usage of coal, oil, natural gas and non-fossil sources (renewables and nuclear combined). For every “building” in the graph the width represented the coal-consumption, the depth oil, the yellow part of the height the natural gas and the green part of the height the usage of non-fossil  sources. But in the end it was only a single row of eight buildings. However, energy usage is not static and AnRep3D allows us to show several rows of buildings, each row representing a single year for all the countries.

So instead of just showing last week’s graph, I collected a lot of additional data, for previous years. We already had 2014, but now 2004 and 1994 were added to cover two jumps of a decade. The last decade (2004 – 2014) covered the financial crisis, where the first one holds the millennium-boom. Interesting periods!

To start with, I will provide you with the original spatial graph through the links below. One is in text, the other is the picture of the graph itself. Remember, we divided the total consumption by the population, to obtain the individual usage for every source of energy. This allows us to compare large and small countries. The total consumption of energy will be the sum of width, depth and total height of the building, so not the volume!

This is the text-link to the original spatial graph on energy-usage per country (remember: WebGL should be enabled). Below a screenshot is presented.

Frontview 8 countries 3 years

This front-view presents the coal-usage per individual (width of the buildings), for the eight countries over three years – the most recent year in front. We already saw France and Italy are hardly using any coal, where the USA and the People’s Republic of China are using a lot. The relative contribution to its total energy-usage is higher for China, but the absolute amount of coal consumed per individual, is more or less the same for China and the USA!

Remember the spatial graph in your browser can be manipulated easily: clicking the right mouse-button, moving the mouse up and down will zoom the graph in and out. Clicking left and moving the mouse will tilt the graph in different directions (or move the observer’s viewpoint around a fixed graph). Double clicking in the graph translates it and moves the centre at the same time. As a result the way the graph tilts will change. Just try it. If you don’t know how to get the normal position back, refresh the graph.

If we fly up a little bit, we can observe the graph from above and in addition to the coal-values (width) now the oil-values  (depth) become clearer.

Tilted view for 8 countries 3 years

The individual oil-consumption is not very different for the inhabitants of EU-countries and the Russian federation. On the other hand,

Topview 8 countries 3 years

we can see people in the USA use about twice this amount individually, while China and India hardly use oil at all. The impression is even clearer with a real top-view.



Side-view 8 countries 3 years

Looking from aside, we can see the individual consumption of natural gas and non-fossil sources (nuclear, solar, wind, hydro and biomass). Now we can also see why France doesn’t need coal. The green part represents a lot of nuclear energy there! Italy doesn’t use a lot of coal either, but the green part is smaller and actually they banned nuclear energy long ago. So the increasing green part comes from other sources and biomass is growing throughout the years in Italy.

By the way: we placed 2014 in the front as the current consumption is more relevant than the historical values. Yet we can see the crisis had its impact and reduced the individual energy-consumption for a lot of countries. Especially the yellow height (so the usage of natural gas) decreased slightly between 2004 and 2014.

In China however, the population grew and therefore the country increased its energy-consumption, but the individual consumption grew as well! The bottom-view shows China’s coal-consumption per individual was more than doubled during the last twenty years, while the USA decreased the individual coal-consumption. Of course it’s an average and in China a lot of the coal will go to e.g. railways and the generation of electricity.

Bottom-view 8 countries 3 years

Interesting is the USA increasing its oil-usage (most likely the result of the shale-oil revolution) and the Russian Federation even more.


Finally, if we compare China to India, we see a huge difference. Although India’s energy-consumption increased, the population grew quicker and as a result the individual consumption decreased a little bit. The individual coal-consumption in India is much lower than in China. For the other fuels, like oil and natural gas the difference is not so impressive, but non-fossil is close (a lot of nuclear energy). Remember: the total usage is the sum of the three dimensions.


With this post we complete the series about energy – abusing our generator. The AnRep3D generator was meant to visualise data from Annual Reports, but other opportunities cannot be ignored!

For people who want to check the individual consumption of coal, oil, natural gas and non-fossils: have a look at my input-file. It was generated by Excel hence all the semicolons. The order is: country, year, natural gas+non-fossil, non-fossil, coal, oil (replacing company, year, revenue, profit, total assets, equity).

Input-file (csv)

The sources used were mainly the for the energy and for the population-sizes. For more information about the AnRep3D-generator itself, visit our website

Posted in Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , , | Leave a comment

Abusing the AnRep3D-generator: visualising energy-sources (III)

My oldest son looked at the graphs and thought it was unfair to combine coal, oil and natural gas + non-fossil in three directions as the volume would suggest a total. Used for the original purpose (visualisation of four different values from an annual report), this objection won’t apply. Total assets and equity are related and provide for a solid base under the revenue and profit, so it made sense to combine them this way. Nobody would think something about the volume – only the shapes of the surfaces have meanings.

If the AnRep3D-generator is used (or abused) for energy, there will be something like “total consumption”, but it is not related to the volume of the building. The total usage is the sum of the three dimensions of the building: depth, width and height.

His second comment was a useful one, because it will allow me to present a different view on energy: countries with a large number of inhabitants will use more energy and therefore the comparison of coal, oil, gas and non-fossil should be “per capita” (by the way: I never understood why it’s not “per caput” as it’s about a single head). Dividing all the values by the number of inhabitants would be an interesting one, because e.g. China uses a lot of coal, but a lot of people live there as well – more than in the European Union and the USA together. We presented only seven countries, so it will be easy to present the average individual consumption for all the fuels. However, I should have added Italy, as it’s a large industrialised country. So this time we will have eight countries, but the amount of Joules shown will be the average per inhabitant. That’s why it’s in GigaJoule (GJ) now, rather than PetaJoule (PJ = 1000,000 GJ). Individuals usually consume less energy than whole countries.

The numbers of inhabitants below are not all exactly for 2014 (as some are ultimo 2015 and some are rounded as well), but it’s only for demonstration purposes and the differences will be quite small:  UK:  65,382,556,  France: 66,759,950,  Germany:  82,175,684,  Italy: 60,665,551, Russian Federation: ‎143,457,000,  China: 1,371,000,000,  India: 1,311,000,000

After dividing all values by the number of inhabitants, the average usage per inhabitant is obtained.This time I used Excel and saved a .csv to be used as an input-file. That’s were all the decimals and semicolons came from. For the meaning of all the numbers, please read the explanation in previous posts (or read the manual in the free demo-package at AnRep3D.)

csv as inputfile

I generated the new graph, but this spatial graph will only be available next time, as we discuss the numbers. At the moment we haven’t uploaded it to our server yet.So for now I put in two screenshots, taken from different angles.

energy-sources per capita

energy per capita tilted

Suddenly China’s individual usage of coal is less impressive if compared to the USA value! The shape of a single building won’t be different but this time the size can be compared as it’s all about the energy-consumption of an average individual in a country.

Remember: the total usage (this time per individual – on average of course) is the sum of all three dimensions of a building: depth, width and height. Width is coal, depth is oil, the yellow height is natural gas and the green height is “non-fossil” (a combination of nuclear energy, wind, solar, hydro and biofuels).

For more details about our generator, able to create spatial graphs or to get the free demo-package: please visit our company’s website: AnRep3D.

Posted in Uncategorized, Visualising Financial Information | Tagged , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , , | Leave a comment