What informers are based on? How the system works, and why i think, that she is revolutionary? Information terminals are more correctly considered in terms of the degree of expansion of existing capabilities., and in terms of improving performance, efficiency and cost optimization.
Used Bloomberg for a while, so there is something to compare. I'll try to tell you my experience with two systems and I'm sure, what will be interesting to many, because. not so many people at the professional level have experience with two systems and are ready to openly share it.
To make the structure of the story clear, I will highlight four aspects
- Completeness of information coverage.
- Ease of access, aggregation and data processing.
- Data visualization and interpretation tools.
- Data collection and loading automation tools.
I do not take all sorts of trade gadgets in view of that, that for me they are not relevant (my trading system is based on QUIK and that's enough). Therefore, initially I will focus only on the need to collect and analyze macroeconomic and financial information. (balance sheets and cash flows).
Rushed! ))
First. Completeness of information coverage.
Certainly, it is impossible to compare the bases directly - they are too huge, therefore, it is difficult to give accurate estimates. But according to subjective impressions after working with two systems, Reuters has a macroeconomic database of about 10, maybe even in 20 times more, than Bloomberg. In other words, Bloomberg is 5-10% from Reuters database - personal assessment.
Reuters, unlike Bloomberg, focuses not only on collecting key data, which are usually shown in the news, but consolidates almost all data, which are, as in open sources, and in closed all over the world! They can be divided into two global subspecies - national sources and databases of external agencies (IMF, World Bank, OECD, UNTIL, The European Commission, from various major universities, banks and research centers). Wherein, if we talk about national sources, it is not only central data, let's say for industrial production or retail sales, and the complete structure in the form, in what it is in the original reports of the statistical agencies. Therefore, the sampling breadth of statistics is so huge., what is infa, starting from the industrial production of slippers in Muhosransk, ending with US GDP data. The coverage and sample depth is incredible, several million macroeconomic indicators for every taste.
Everything is much simpler in Bloomberg. I remember trying to find the structure of the debt burden of the population and households by 10-15 the largest countries in the world. Not to tell, that the request is out of the ordinary. But there was nothing. Seems, something in the USA could be ripped out, but find the structure of assets and liabilities by economic agents (population, companies, state, banks) 15 countries was difficult, if at all this data is there.
In Reuters, everything is presented in full. It is impossible to list everything. Over the past month about 40 spent hours just, to assess the boundaries of coverage, understand, what kind of data is presented there, but about a third advanced. It's a whole universe, where you can find everything, anything. You can dig there for a long time, but it can be very interesting, because. I learned about the existence of most of the data for the first time.
In Bloomberg, it is easier to list what is on the macroeconomic information, and in Reuters, on the contrary - which is not. From the point of view of macroanalysis, I Bloomberg hardly used, because. the sample was rather scarce and was not suitable for structural analysis. Data could still be found across the United States, but in other countries a little. There is so much information in Datastream, that everything is limited only to the flight of thoughts and ideas…
Second. Ease of access, aggregation and data processing.
There are tens of thousands of different data sources in the world, which have their own specifics and presentation format. In American and European statistics, it is quite convenient to collect data manually., what can't you say, for example, about chinese statistics. Data is too fragmented, no uniform presentation format. Reuters systematizes tens of thousands of sources into a single whole, in a single standardized format.
Datastream's Biggest and Most Fundamental Advantage over Bloomberg in an Ultra-Easy Data Navigator.
Here's an example.
In this structure, data for all countries. If you need information on monetary aggregates, then, to switch from the United States to Korea or any other country, it will take no more 2-3 seconds! Any mood indicators, national accounts in full structure (by expenses, income, added value), monetary and banking statistics, prices, labor market and employment, industrial sector, consumer, public sector and so on. Basically, comprehensive presentation.
Therefore, there is never any problem finding the information you need., because. with a concretized structure, navigation is very fast and you always understand, what can be found in the section.
There is no such thing in Bloomberg. There are templates by country, where key data is collected, but very little data, so search is usually used. But you need to know exactly under what name the indicator is presented in the database., otherwise you will have to swarm around in the garbage for a long time and manually filter.
A simple example, to make it clear. Once I tried to pull out the G20 GDP structure from Bloomberg for loading into Excel. We can say the simplest request possible. This operation continued for me. 3 hours somewhere and ended on 5 or 6 country. It was possible to get it of course, but at about the same speed, as if doing it manually through open national sources. For this reason, I did not use Bloomberg to collect data on macro indicators., because. easier to shoot, than looking for info there. Individual indicators can be found, but to compare several countries and have a global picture before your eyes? Not, there is no such thing. Everything is rather primitive in this aspect..
I did a similar task at Reuters and it took 10-15 minutes about! If you train, then you can do in a few seconds. OK, this is the simplest, and if something is more complicated? For example, structure of assets and liabilities of the population and companies by countries of the world? In Bloomberg, I dare to suggest such data, in principle, does not exist, and on the first try in Reuters it can take about an hour or two, but if you know where to look and create an automated script- sample, then less 5 minutes.
By key data, allowable, industrial production, inflation, money supply, trade balance, GDP (no structure) on 25-30 countries data are selected for 3-5 seconds, whereas in Bloomberg it would take a day or two, and manually a week.
Datastream's scalability and performance is endless. It depends on how wide you look. Compared to Bloomberg, then the performance improvement can be from 3-5 once before 3000 once. That, what can be done in Reuters for 5 seconds in bloomberg might go away 10 hours.
This was the biggest discovery for me in the terminal., because. I have worked with various data for almost my entire adult life (trading, financial, macroeconomic and other), I know how much time and effort it takes, to find something. When the collection of a certain base, which I borrowed 3 hours now runs in 3 seconds - yes, the first time, such productivity gains are shocking, because. you realize, how limitless the potential of the system. In terms of productivity growth, this is definitely a breakthrough. That, what did entire scientific centers with a staff in 500 person per month, now I can do it alone in just an hour - productivity growth in 88 thousand times (500 * 8 hours * 22 of the day). Another thing, how to apply this performance correctly? If there is no skill and experience, it's like flying rocket engines over a puddle ))
Another analogue for working with macro data in Bloomberg and Reuters. It's like going to St. Petersburg from Moscow on foot or by plane, train. Can, certainly, walk sooner or later, but on a plane or train it is more convenient and faster like that?
Datastream is a unique aggregation system, integration, structuring and systematizing huge amounts of unstructured information. Faster, what to collect there, I guess, nowhere impossible and not sure, that in general someone will create something similar.
Third. Data visualization and interpretation tools.
For example, no need to download data, but I just want to see the dynamics immediately and quickly…
Usually everyone has problems with this.. Yes, there is a lot in the public domain, but for now you will find, while you download the data, To view. Good, if the website of the statistical department has visualization tools, but usually everything works inconveniently, taking a long time. In terms of one data series, this is, at best, about a minute.. You won't see much.
Datastream has an ingenious preview tool. Genius in its simplicity and speed. Works mega fast - less than half a second. Move the cursor over the indicator and without opening additional pages, without loading data, a small window appears with a graph and an explanation, where the data comes from, what it is, how often is it updated, etc.
That's the way
This is asset data. FED behind 63 year from Z1 report
There you can calculate the dynamics year-on-year or quarter-on-quarter., if data is needed, then download and so on.
Bloomberg doesn't have it. Although, when I used it I really wanted to have such a thing, where could it be for 2-3 minutes to view the whole pool of indicators at once (about 100 things) to assess the operational tactical situation.
But Datastream is equally valuable in an even cooler tool, как Datastream Charting. He is so good, that I even gave up Excel and data loading, because. this thing surpasses Excel in many respects. Have you seen, that the last month almost all charts in my articles from Datastream Charting for one reason - fast and convenient.
Visualization tools? This is how the graph looks with default settings.
But if you finish the design, then you can do this.
In terms of design, everything, whatever - colors, background, fonts, dimensions, graph and border style, transparency, all sorts of markers and so on - a wide range, more even, than in Excel.
But not only this is so good Datastream Charting
There is something, what is lacking in many terminals, which makes it necessary to load data into Excel for primary calculations.
Datastream Charting has a powerful math module for calculations (feature set is huge, so I will highlight the main)
- Percentage with the selected period (kV, year or some kind).
- Absolute relationship with the period (kV, year or your)
- Amount for the period (for that, to calculate annual results for quarterly data, to period 4, and if for menstruation, then 12 period)
- Accumulated amount (useful for cash flows)
- Average values for the period (for example, draws daily chart S&P and calculate the average values for the week, month or quarter)
- Anti-aliasing (simple and exponential)
- Trends (current and forecast) different techniques
- Logic functions
- Math functions (ABS, EXP, LOG, LN, POW)
- Statistical (huge set) and naturally there are correlations. You can do regressions
- Technical analysis tools
- You can also make lags (shift the indicator on charts in comparison with another), to create all sorts of leading indicators.
- Well, fold the standard set (for data consolidation), subtract (for spreads), cleverness (for example for converting into currency), divide (for relationship, share of something)
- Fix rows to point, to depict the comparative dynamics of indicators, ie. when we equate to 100 and calculate the percentage change from a fixed point on the selected date.
- Automatic conversion to the selected currency. Also useful functionality, if cross-country comparisons are made
There is nothing like this in Bloomberg, although it was also sorely lacking. Generally speaking, that Bloomberg is not very suitable for working with data and charting, no matter how strange it sounds. Not, well, the eurobucks chart versus S&P can be done, gash the US GDP there, technical analysis to stick, but this is not the right steppe at all. I have not been able to adapt Bloomberg to work with macroeconomic data in their system., not convenient just. Download some data on China, which you can find yourself too lazy in national sources to look for in Bloom, but to create full-fledged databases for analysis – not so efficient and fast, as in Datastream. Bloom – it is rather a system for short-term conditions, but not for analyzing and working with data.
Any forex cuisine presents tools of technical analysis and graphical analysis, there is nothing valuable and useful here. It is the mathematical module that is needed to work with data series, so as not to load in Excel, kind of built in, automated analogue of Excel. Load the good old Excel for more complex and large-scale research. Datastream does it like a masterpiece. If i knew, that Reuters has such a powerful system, I would have moved earlier.
Fourth. Data collection and loading automation tools.
The last item left for a snack. My old dream was to create a system, which would update the data of interest to me in automatic mode. So that the temporary investment in creating a template is only one time, and then the system starts working. Previously, I had to manually finish, update databases, to keep the info up to date. I had to discard all unnecessary, leaving only the highest priority data is not bad. But even they were too lazy to update every time - this is a routine process., like walking in circles, especially since there is a lot of data, you can't keep track of everyone. Ideally, I wanted to eliminate the routine, leaving energy and time for creativity and concepts.
Infographics, which you have already seen - this is the template, only shared on the network. According to any data, you can do this. So far on government bonds, but you can, as for macrostatistics of any format (primary or processed data in calculations), and according to corporate statistics. Can be done, so that company reports are updated and presented in the form of a graph! Now everything is limited to fantasy and desire. Datastream is definitely a revolution in many ways, which changes the understanding of working with data.
In addition to working in the system, there is also the possibility of loading data into Excel and intuitive programming of templates. The scope for creativity is very significant. All companies in the USA can be selected, who have more than 1 billion dollars and calculate revenues on them, profit, assets and half a thousand other indicators in the entire history of publication of corporate reports (there are rows with 1985 of the year). And you can download the structure of GDP for the G20 for 60 years or Central Bank assets from the G8 - yes, all, anything. Or create a bond or stock price monitor in this style.
I wrote scripts for importing data into Excel for quick monitoring of the current situation.. I made the same matrix for currency, chest of drawers, stock indices and desired stocks. I also think in Excel % change to period. Change in assets by week. back, month, quarter, year and on the selected date. Everything is immediately visible on one sheet.
By the way, only for unloading corporate reports I used bloomberg )) Bloomberg has some templates too, but the time costs for their programming are incomparable. Already tried to do there and did not like it, because. convenience leaves much to be desired. Hell itself is to search and carry codes, tickers to download to Excel, because. hard to find and take a long time to pull in Excel. By companies it is fast and in an instant, but according to macro indicators, forever. Everything is done so quickly at Datastream, that within the accuracy of counting for any data. You can select at least a thousand indicators less, in less than a minute.
In Datastream, through an advanced filtering and search system, you can do all this in a few seconds or minutes, depending on the size of the database.. Comfortable!
Another convenient functionality for preparing automated reports. If templates are created, then you can load them all at once or in sections in Word, Excel or as screenshots.
Datastream – this is technology, that are changing the world of statistics and working with data. I have never seen anything more convenient and the mistake is, what I learned only recently.
With great surprise I discovered, that people are completely unaware, what opportunities exist, even among those, who is professionally represented in the financial markets. I used to think, that Bloomberg is the pinnacle of civilization, and Reuters is a news provider and some kind of incomprehensible database of financial indicators, like the price of wheat or oil, which are of little interest. But the last month was a discovery for me. Now I got it, that Bloomberg is far behind in macroeconomic statistics and data. It is foolish to hide such opportunities from readers., therefore made a comparative review.
Since several people have already asked me, where are the charts from and what is Datastream, I'll write a contact right away, so as not to answer questions of the same type by mail. On the system, write not to me (I am not a Reuters representative), and on this contact
[email protected]
I have only considered the macroeconomic module so far.. There are also corporate statistics on the countries of the world. (more 30 thousand companies) with full financial results for all available periods, as per direct reports, and in standardized form, so that you can compare companies from different countries. Company debt data, structure of bond issue. Shareholder data (shareholder structure), on transactions from mergers and acquisitions, all sorts of corporate events, analysis and analytics, forecasts (for those, who needs them), many models for evaluating financial indicators, convenient comparison with competitors and much more. In Bloomberg at the level, but the base there seems to be smaller, because in Russia in Bloomberg fewer companies, than in Reuters.
On the debt market, a database of all issued bonds by all issuers (companies, state, state funds). You can calculate the distribution of redemptions, as I did by country and dozens of other schemes and models. Bloomberg did not look at this aspect.
For investment banking. IPO type, SPO, M&A and others. Maybe, the only thing, what is better implemented in Bloomberg.
All kinds of trading opportunities, option calculators, debt market, CDS and stuff it's from another opera. Haven't looked at any Bloomberg yet, nor in Reuters, because. my trading system is now integrated in the Russian market and their trading potential is not particularly interesting.
Useful thing, like tracking sheet. You drive into the sheet, allowable, portfolio shares. The system keeps track of all the selected shares (corporate events calendar, past events, all processes, news, deals and so on, statements and comments of company representatives, plus the financial reports themselves and you can aggregate them, ie. consider an automaton p / e, p / s for portfolio and a hundred other indicators, comparing with sectors, industries or indices) Useful in that sense, that you will not miss important events, well, and the corresponding concept of alerts for events.
That's it, if you make an integrated assessment of the capabilities of systems from the point of view of the macroeconomic sphere, then Datastream is unambiguous 10 from 10 (masterpiece!), and Bloomberg 3 from 10 (almost like working manually or easier through aggregated bases from the IMF and OECD)
According to corporate statistics, they are approximately equal. Bloom mainly used for corporate reports. In real time, financial instruments are also equal, but not Datastream right there, а Eikon, including for various calculators and plugins for trading. For the rest, I did not go into details.
I can tell, I have found, what was looking for ) There is still an advanced programming system in Excel, where many ideas can be realized.
author:
