1st Quarter 2015, Issue 35
In this issue, we welcome Frost Consulting as an Associate Member, we highlight an article contributed by ANALEC, we get an update from our Standards Committee highlighting our objectives for 2015, we highlight the business drivers for Componentization, we present the case for “Why do you need tags?”, and how RIXML is applied to research, from RIXML’s Implementation Guide, we promote Marketing & Membership and we get a perspective from Jack Roehrig, Executive Director of RIXML.org.
RIXML.org, a consortium of buy-side, sell-side and vendor firms committed to the development and implementation of the first open standard for investment research, is pleased to announce that Frost Consulting has joined RIXML.org as an Associate member.
Frost Consulting is a specialist advisor in investment research procurement, management and distribution. Frost assists research consumers (asset managers/plan sponsors/SWFs), research producers and associated service and technology providers to leverage regulatory and technological change.
"RIXML.org welcomes Frost Consulting as an Associate member," said Jack Roehrig, Executive Director of RIXML.org. "Frost’s knowledge of, and participation in, the industry discussion around regulatory change related to research, combined with their expertise in creating client solutions, will bring valuable perspective to RIXML.org’s ongoing mission."
"Frost Consulting is pleased to be partnering with RIXML.org to help create forward-looking industry solutions for research consumers and producers", said Susan Walton, Principal at Frost Consulting.
About Frost Consulting
Frost Consulting is a leading authority on the global investment research procurement value chain and related market structure/regulatory change. Frost assists a wide variety of clients in developing commercial and operational strategies to maximize research ROI/alpha generation, optimize distribution and mitigate risk in a rapidly changing commercial environment. www.frostconsulting.co.uk
Turning the traditional model on its head
Every analyst has a fixed amount of time to produce research and call clients. However, this traditional approach works on the premise that once I have my report, I need to sell it. What if the selling was taken care of by increasing the presence of the analyst on the mediums by which everyone has shifted to? You receive information through WhatsApp, Facebook, Twitter, LinkedIn and news feeds where you as a consumer switch on and off rather than being constantly bombarded. Yet the model for research is largely unchanged.
Making it easy for the analyst
It is all very well to tell the analyst that he/she must become more commercial but isn’t there any way that technologists can help them? RIXML is the standard for information interchange, but as we are all aware, developing feed mechanisms to transmit information from one system to another is nothing new. The key however, is that we remove the tedious technological requirements and focus on the business. This is the power of the standards, notably RIXML.
To provide the analyst with the framework to take their existing working process and capture and componentize their work without them really knowing is the Holy Grail in terms of the next stage of research. Systems already exist where authors can Tweet, use Facebook, LinkedIn and news wire information, safely within the confines of compliance and regulatory approval with infinite possibilities for the analyst to be more commercial without adding to workload.
RIXML Componentization – The tool for the job
Through componentization, the authoring of a report is now a beautifully constructed dissected creation. There is a place in RIXML componentization for every consumer and system. The quant feeds his models instantaneously. The signal-only consumer gets the lowest level of the report - i.e. buy X, sell Y. The on-the-go buy side analyst gets a 20 word summary with more detail as required. And the diehard consumer gets all the financial information and detail delivered in the way desired. The analyst does nothing different than they did before – one report (browser, Microsoft Word, PowerPoint, Outlook – irrelevant, they are just tools).
Having used componentization for many years, we all know that it works. Taking a detailed report from the institutional side of the business, converting it to a retail banking report, a presentation pack, tweeting out the comment to the community, and giving the meat to downstream systems in a standard way is what we technologists can do to help the analysts. RIXML’s continual evolution is a function of driving new norms to an industry that is slowly catching up with the times. Through the RIXML organization, we should be able to provide all the correct tools to allow the evolution of the industry to occur as more and more institutions bring research creation and delivery in line with the new world order.
Are you trending?
One of the most exciting areas of RIXML is the inclusion of linkbacks as a working enhancement of the standard. At the moment, linkbacks are virtually non-existent in many 3rd party contributors and are so important to establish the voice of the customer.
Take a look at any social media, website or mobile app and you will see the ability to like, share, dislike. These concepts are quick, simple and convey a wealth of information when aggregated across all ecosystems. The term ‘trending’ again has become part of our everyday lives. Can we not assess whether a report in whatever shape or form it is being consumed, is trending? As the linkback standard moves through its versioning, the inclusion of fundamental concepts will be a great asset as technology providers, institutions and consumers try to make sense of, and repair the broken business model.
Consider the analyst producing their 20 page masterpiece on a particular company or sector. Traditionally, the salesforce will take the idea and sell to the client base. The client can either agree or disagree with the view. Even with the best feedback mechanisms the client base that gives their opinion is still very small because of the medium in which the information is captured (usually as a result of a phone call or someone reading an email).
Enter again the notion of linkbacks and social media. Collect enough of the information on how many times a report was read, whether it was liked across a massive base (or at least much larger than the traditional one-on-one calling) and you can start to drive meaningful stats. It also allows for the ability to assess quality of output from the analysts, the way in which this output directly affects stock pricing through the many layers of idea, liking, action, flow and effect.
‘RIXML Events’ – A new piece of the big data puzzle
Other factors outside the control of the analyst are the macro factors which can affect a whole industry or sector. At the moment there is no unified way of conveying this to the consumer base. The RIXML standard is the prime candidate to deliver this information. For example, oil pricing or a breakout of hostilities. What should stop an analyst getting out a short tweet, message or idea on the effects of the changing oil price or what the outbreak of hostilities will do for the oil sector? Consider a RIXML structure driven by events and ramifications; it is all possible under the structure. That event can also be liked, disliked and shared. Does a particular consumer of the event agree or disagree with the outcome within the particular industry vertical he/she covers. The tools are already in place to do all of this. Get this level of granularity from the consumer and the big data value add speaks for itself.
2015 Objectives for the RIXML Organization
Looking forward into this new year, the RIXML organization had several discussions about topics of interest. We sought to create a list of objectives to guide us in future efforts.
Componentization
Finalize and productize our documentation detailing the guidelines for componentization agreed by our Working Group. Complete a pilot program to illustrate and exercise our ideas. Eventually we will move to v2 ideas, such as inline tagging (below the section level).
Social Media
Propose specific modifications to the RIXML schema to facilitate the inclusion of social media messages within both new and existing Research authoring and publishing platforms. Pursue the RIXML organization’s own social media plan toward better engagement with interested parties on common social media platforms.
Link-Back Landscape
Dig into the timely subject of link-backs. With many research publishers broadening their platforms to embrace various forms of digital content delivery, issues around Identity & Access Management (IAM) arise between publishers, consumers, and aggregators. RIXML should consider what value it might add in this space. Perhaps a technical white paper would be useful to ensure implementation standardization. (Note that this is a distinct topic from prior discussions of entitlements.)
Big Data
Continue to monitor the opportunities for RIXML at the intersection of the Investment Research marketplace and the application of “Big Data” methods toward discovering actionable investment signals. Michael Mayhew and Gabriel Lowy have shared their insights on this topic with us in the past. Perhaps we can extend those relationships.
Side-Car Schemas
Explore the adoption of these schemas and evaluate meaningful updates and additions. (The release of RIXML schema version 2.4 in 2013 included a pair of “side-car” schemas intended for communicating analyst roster and coverage universe data.)
Spot Tags
Propose one or more specific solutions to address the need to avoid fragmentation of keywords in “breaking news” situations. RIXML should offer an easy method for research content publishers to tag new products with non-canonical keywords in a fashion consistent across publishers and in timeframes much shorter that the RIXML schema release cycle.
Emerging Technology Committee
Richard Brandt led three calls since our last Quarterly Meeting – on October 3rd, November 7th, and January 9th. Topics of discussion mostly followed the activities of our two active working groups – Componentization and Social Media. In addition to the regular calls, the group met for an in-person workshop on December 3rd at Jordan & Jordan in New York.
Componentization Working Group
Consistent with our overall plan for Componentization, at this Quarterly Meeting we are making available a draft of the guidance document. It incorporates all work to date from calls, workshops, and the Wiki. Please review it and submit any feedback to our program office or to any group leader. After the content has been reviewed and finalized, we will productize the document and post to our web site.
The next key step is the formation of a pilot program to put the guidance to work with specific examples. This will provide some grounding to what we expect to achieve, and bring greater clarity. We’ll need to define both a data context and a workflow context at the aggregator level. We’ll need to determine how HTML5-formatted components are moved from place to place. And we’ll need to consider how components work in a link-back approach vs. a content feed. The pilot program should also provide results that lead into an eSeminar to help spread awareness.
Social Media Working Group
We will put together a straw man on how the schema might be extended to incorporate content intended for social media platforms.
The business drivers around monetization of content are getting stronger. The timing is right for RIXML to issue guidance.
> Decreasing commissions, creating a greater sense of urgency for research discoverability and payment realization.
> Evolving technologies causing transitional time for the research business to update delivery models.
> Increasing buy-side clients’ use of mobile devices and social networks.
> Changing channels for packaging and delivering proprietary content.
> Transitioning from document-centric toward component (unbundled) delivery models.
> Growing presence of non-traditional players in the marketplace.
In the coming weeks, we will be issuing a DRAFT release of our Componentization Guide for comment---stay tuned. Componentization remains a core topic for our technology working groups to address throughout 2015.
Imagine that you make and sell a large number of products, each of which has hundreds of pieces. To package these products, you have three choices:
- You can put all the pieces for each product into an unlabeled cardboard box. From the outside, every box looks exactly the same.
- You can put all the pieces for each product into a cardboard box that has a label on it giving a basic description of what it contains.
- You can put the pieces into multiple boxes that organize the various components, each with a very specific label describing the contents. For the more complicated sections, there are boxes within the boxes to keep everything well organized. All of these boxes are nested neatly in one big box, which has a very clear label describing the contents.
Which would you choose? Which one do you think your potential customers would prefer? In this example, the packaging makes a huge difference in the usability of the product itself. The product can exist without good packaging, but it becomes far more usable – and valuable – with it.
What is this product? Investment research. What are the pieces? All of the valuable information – company information, financial data, recommendation information, and analysis – that each report contains.
XML is sophisticated packaging for electronic information. Furthermore, RIXML is sophisticated packaging customized specifically for investment research.
To apply RIXML to a particular piece of research content (such as a research report or an audio file of a morning call), the relevant tags along with the information that is described by these tags are formed into a tagging file. This tagging file is called an instance document. This instance document does not replace the research report or audio file. Instead, it travels along with the research item, providing the necessary tagging to the desired data repositories. Within the instance document are the tags and content that describes the content in the research item (for example, ) as well as the tags and content being used to uniquely identify the research item (for example, Strategy_Daily_20120512.pdf). By standardizing a common set of tags for the information that is generally included in research content, such as security identifiers, industries and sectors, ratings and recommendations, asset types, etc., the creators, aggregators, and consumers of investment research are able to vastly improve the process for getting the right research into the hands of the consumers who want and need it. In addition, this advanced level of tagging and access to individual contextual elements allows for a new generation of alerting mechanisms, navigational frameworks and enhanced user interfaces, and delivery to mobile devices. The remainder of this Implementation Guide contains details on implementing RIXML, core concepts behind RIXML, and details on the schema itself. These sections provide information that is critical to understand in order to begin to leverage the strengths of RIXML.
Interested in joining RIXML.org? Call our Program Office and 212-652-4470 or email us at to=This email address is being protected from spambots. You need JavaScript enabled to view it." target="_blank">This email address is being protected from spambots. You need JavaScript enabled to view it. for additional information.
As we build out our Events calendar, we welcome direct participation in members’ events, as part of panel discussion, topical roundtables and e-Seminars/Webinars. Related, we also encourage members’ direct contributions to our recently revamped web site(www.rixml.org), through the additions of whitepapers, RIXML.org experiences/forums, in the spirit of promoting/adopting the schema and best practices.
ALL of our members are part of the Marketing & Membership committee !
PLEASE NOTE: This viewpoint is entirely my own and neither the official viewpoint of RIXML.org nor the viewpoint of any of its member organizations.
“Cosmic” Research: Mr. Spock, A.I., Big Data, Carl Sagan and Yotta, Yotta, Yotta
Now that there are very early signs of Spring and life emerging here in the Northeast(the Witch Hazel and Snowdrops/Galanthus are starting to bloom in my yard) after a very harsh winter, since Spring brings hope and hope springs eternal, it is always fun to think of what is possible – and what lies ahead that may be new and exciting – Spring always give us that opportunity.
Is seems fitting, when examining the “cosmic” opportunities that may present themselves in the Research space down the road, to remember one of the ultimate all-time cosmic characters, Star Trek’s Mr. Spock, R.I.P, who was portrayed by Leonard Nimoy, who passed away(the final frontier ?) at the age of 83 on March 2nd. Mr. Spock, a logic-driven, half-human science officer of the starship Enterprise, is an iconic figure to generations of “Trekkies”, who advised us to “live long and prosper” – a noble sentiment, but, we are not sure what investment time horizon Spock had in mind or what his risk tolerance was, but, I digress.
From a different “cosmic” perspective, 2015 seems to be shedding new light on the roles that Big Data, combined with re-emerging Artificial Intelligence tools, may play in the Research/Investment Analysis space, to take us “where no man has gone before.”
Firms like Kensho have emerged, which is out to do for financial analysis what Google did for search. A Big Data article published in Forbes last May highlights Kensho’s software, dubbed Warren(as in Buffet), that offers up a simple Google-like text box, in which you can pose very complex questions – all in plain English – such as ‘Which cement stocks go up most when a category 3 hurricane hits Florida ?”, or, “Which Apple suppliers’ shares go up the most when a new iPad is released?” The article mentions that out of the box Warren can find the answers to more than 65 million question combinations in an instant(might IBM’s/Jeopardy’s “Watson” be jealous?) by scanning more than 90,000 actions such as drug approvals, economic reports, policy changes, political events and their impact on nearly every financial asset on the planet – COSMIC ! As the black-boxes look to seek and find their right level in the space, the question and discussion of data quality(garbage in/garbage out ?) remains.
Gabe Lowy, who in the past has kindly presented his thoughts to RIXML.org members, posted in his February 24, 2015 Tech-Tonics blog “Data Quality Determines Decision Outcomes”, that Big Data does not change the relationship between data quality and decision outcomes – it underscores it. It is in the integration and management to provide the highest quality data in the timeliest fashion at the point of decision – regardless of whether the decision maker is an employee or a customer. Lowy emphasizes that, as more companies seek to gain deeper insights to improve operational performance and competitiveness, greater attention needs to be paid to data quality and the validation of models built on that data.
We know Big Data is BIG(MORE Data – Warp Speed !) and getting bigger - even the size and scale is hard to comprehend – I came across a relatively new entity known as a Yottabyte. A Yottabyte is a measure of theoretical storage capacity and is 2 to the 80th power bytes, or, in decimal, approximately a thousand zettabytes, a trillion terabytes or a million trillion megabytes – COSMIC ! Millions and trillions – I am hearing Carl Sagan’s voice for some reason. The notion of portable Yottabyte devices down the road may not be far-fetched…impressive for those of us that happily bought the original IBM PC with an astounding 256K of memory – Beam us up Scotty !
Today, the press and social media are awash with thoughts on A.I. with its potential perils and the inevitable question, “What’s different this time?” Matt Turck’s January 5th blog post mentions the resurgence of A.I. is partly a child of Big Data, as better algorithms(in particular what’s known as “deep learning” pioneered by LeCun and others) have been enabled by larger than ever datasets and the ability to process those datasets at scale at reasonable costs.
While early, RIXML.org member organizations are examining this space more closely – to the extent “FAQs” are brought to RIXML.org working groups to educate all members, that is a good thing.
As for understanding the relevance of these “cosmic” developments to the more grounded issues of the day, i.e. the timing of the first blooms of the Witch Hazels and Snowdrops, that is also a good thing.
Live long and prosper…….
Gustav Mahler