Monitor110 and Me - A Big Week

Let me start by saying that Information Arbitrage is my personal blog. I write about what interests me, specifically those things that are "burning a hole in my head" that just need to find a way to get out. I have found tremendous freedom, satisfaction and comraderie in blogging - I can't believe that it has only been a little more than two months! Thanks to all my readers, commenters and those who have reached out to me personally. I can't tell you how much I appreciate your becoming a part of the online dialogue that has, in many cases, become an offline dialogue due to shared interests and ideas for collaboration. Whether you have agreed with me or not is not the point; it is your engagement and passion that matters. You all rock.

I do other things besides blog, however. I do selective angel investing both individually and with my partner, Jeff Stewart, through Geometric Group. I also invest in hedge funds. But principally, beyond being a devoted husband, father and friend, I am President and Chief Operating Officer of Monitor110, Inc. This is a role I love and the passion I chose to pursue after leaving Wall Street. It has been a magical 18 months, working with my partner Jeff to build the team, the technology, and the platform to make our shared vision of "Helping Institutional Investors turn Internet information into alpha generation" a reality. We recently came out of "stealth" mode, announcing our financing partners in a press release a few weeks ago. Reflecting on all that has happened since Jeff and I met at the beginning of 2005, I have neither felt so fortunate in my business career nor been so proud to be part of such an amazing group of people. Smart, dedicated, multi-dimensional, and, most importantly, fiercely passionate about our shared vision. In my spirit of blogging, e.g., the things burning a hole in my head, I feel compelled to discuss a little bit of what went on this week with Monitor110.

For starters, our company was discussed on the Front Page of last Thursday's Financial Times, as well as being featured in a detailed write-up that same day in TechCrunch. The FT story actually hit the online edition late Wednesday night, and was quickly picked up by Drudge, Techmeme and many top sites during Wednesday evening and into Thursday. Suffice it to say, this is pretty dizzying stuff. The number of hits on both the Monitor110 website as well as my blog was, at least for us, quite impressive. My blog also got some play around this same time because of comments I had made regarding the Amaranth debacle, which were ultimately linked to by respected finance bloggers such as Trader Mike and Howard Lindzon (both of whom also linked to the FT story about Monitor110), as well as being carried on both Reuters and NPR. I have to say thanks to all of those who reached out to me in the wake of all this exciting news. The well-wishers and excited friends from the past, as well as new friends from the blogger and technology communities were awesome to behold. Whew! What a week...

So a little bit about Monitor110. As some of you know, Monitor110 is a company that is working to address some of the passions and possibilities I have been yearning for; namely, a tool for helping institutional investors access, analyze and monetize information from the internet, as a powerful extension of what they are getting from their principal news and market data portal (i.e., Bloomberg, Reuters, and Thomson). Why have we done this? Our fervent belief is that the internet is the next great data set to be monitored, mined and monetized, as has been the case with ticker plant data and what I'll call "Tier 1" news data. Reuters ruled the day in the late 19th/early 20th centuries by leveraging technology - undersea cables (and before that, carrier pigeons) - to transmit news across the globe. This gave them a powerful advantage that was cemented over decades. Bloomberg came along in the early 1980s by giving traders (initially fixed income traders) access to data and tools - utilizing client server technology - that were previously unavailable. They have created one of the most powerful information services empires of all time. We think the opportunity presented by the internet - by leveraging the latest in text processing, signal intercept, machine learning, memory and storage technologies - has the opportunity to be the next great wave in the Information Revolution. And the power of these technologies can and will be harnessed for the institutional investor, and Monitor110 expects to be among the vanguard of this revolution.

Let me say upfront - we've bitten off a lot. As we all know, information from the ever-growing body of the internet comes in a variety of forms and from a variety of sources. These sources can range from Nobel Laureates to stock spammers, from accomplished scientists to shills, from thought-leaders to kids discussing Pokemon. This is why simply having a "fat pipe" is a necessary but insufficient criterion for delivering a useful and powerful tool for institutional investors (or anyone else for that matter). Our clients need data, they need it fast and they need the stuff they're interested in and for it to be easily discoverable. Ok, so how does this happen?

Clients need data. Right. We have focused our efforts on what we call the "dynamic" portion of the internet - those places where content changes on a regular basis. This includes blogs, company websites, Government sites, regulatory sites, union sites and alternative news sites, among many others. So this data set, while a fraction of the size of the Google archive (which amounts to billions of static web pages), is currenly in the tens of millions of sites and pages, and is expected to reach 100 million by the end of next year.

They need it fast. For institutional investors it is all about real-time. So, what you need is the intersection of a ticker plant with the internet. This requires pretty slick architecture together with massive processing power. Check. We've got that.

They need the stuff they're interested in and for it to be easily discoverable. Whoa. Here's where it gets really hairy. The stuff they're interested in...Hmm...That sounds an awful bit like, what's that called, relevance. So, as each piece of text flows through the system, it is given a score based upon its relevance to the query (which is generally either a company or an investment theme our client is interested in). Ok, check. But, as mentioned previously, how do I know that the piece of information in question isn't from a spammer, a fraudster or a charlatan of some stripe or variety? I know there is a word for that, what is it....Right - reputation. So, by using proprietary algorithms to analyze the reputation of the site from which the text is derived the Monitor110 system can help the client decide which of the relevant items it wants to analyze. So, we put relevance and reputation in the control of our clients, giving the them the ability to either shrink or expand the data set based upon the types of data they are looking for and the nature of the online dialogue taking place in a given company or investment theme.

There is one key thing I have neglected to mention. When building systems to extract meaning from data there is one thing above all others that can help increase the value of the system's output. What is that thing? I'll tell you - context. What this really means is domain knowledge, or, in our case, using the language, structure, and taxonomy of finance and investment to train our systems (read: persistent vertical search tool for investment) to assign greater weight to certain types of text. When you do a search in Google, does it know you are an investor looking for investment-related documents? No. As an investor, do you particularly care what the most popular return to a Google search is or do you just want to find what it is you want to find? And, when setting up a persistent search (one that gives real-time notification when relevant documents are identified), are you worried about spam clogging your alerts, or is such a document scored as being low in reputation and, therefore, to be ignored or completely filtered out?

Long story short, we believe that targeted vertical search for areas like finance and investment represent the state-of-the-art and requires the intersection of leading-edge technology and domain expertise. These are the skills resident within the Monitor110 team. Helping institutional investors make money by being on the left hand side of the New Information Dissemination Cycle - this is what we are all about. It's not about being afraid of the crap that we all know is out there - its about developing the tools to identify the good stuff, quickly, and making it readily discoverable for our clients. We are still in beta. But we're almost there. It's getting really, really exciting.

Stay posted for some recent examples of our tool in action. It's pretty cool.

Comments

[Roger]

John, thanks for your understanding and your comments certainly reflect a plethora of domain knowledge. It sounds like our backgrounds are quite similar. And thanks for clarifying your position on the social aspect of data use. I don't think we fundamentally disagree on long-term value of such data, only how it is best introduced to and used and valued by subscribers. We'll see how it goes, but yours is certainly a valid point.

[John Wilson]

Roger, I can understanding your nervousness about answering my questions.

Based on your response, am I right in concluding that you are operating a multi-tenanted solution but each customer is isolated from the activities of other users i.e. no collective intelligence, nor aggregation of inputs across the user base. Will you be offering a on-site installation version?

I'm very familiar with the client base to which you refer having worked with/in many of large institutional buyside and hedge fund business (as well as having been on the sell side, with time at Goldman Sachs and on the Board at Man Financial). So I can understand their initial preference to silo their information. However, I'm not suggesting any "social" aspect be introduced on trendy web2.0 grounds or the like. I remain of the belief that by participating in the pooling of data, firms will gain additional market insight ("crowd noise"), achievable through anonymous data aggregation.

[Roger]

Yaser and John, thanks for your questions. They are both intelligent and relevant. However, we believe our development roadmap and marketing strategy represent a competitive advantage, therefore we make it a point not to discuss our development roadmap or marketing strategy. That said, suffice it to say that many features of interest to institutional investors are either being incorporated into our Release 1.0 tool or on our roadmap, the prioritization of which will be driven by our clients.

One thing I will say is that we are not currently contemplating a "social" aspect to our system. The clients we are working with, those whom I know from my years on the Street and those in whom I'm personally invested have no interest in sharing their intelligence or behavior on a systematic basis with others. If they could find this stuff out about their peers, great, but are not interested in sharing themselves.

It's just not the culture of professional institutional investment in general, or hedge funds in particular. I know this is a generalization but at least among our initial target market I know this to be the case. We are focused on what is happening out on the internet that is relevant for our clients, not specifically what is happening across our user base.

[John Wilson]

As I understand it Monitor 110 will be providing real-time; relevant; trustworthy summaries on textual inputs. A few questions occur to me.
- Am I right that time series tick-data is not included as an input in the current product capabilities?
- How often you will be "polling" relevant blogs for new posts, given that there presumably is a time value associated with the output as well as an information value?
- Will the addition of new sources of data selected by users (which I'm speculating users will be able to do) be available to only the person that adds them or passively for all users of the system to include in their scans?
- Will I be able to perform a reverse scan of how many people (not necessarily who) are monitoring which sources and what is being monitored (perhaps with top 10 list etc)?
- Who will have access to observe who is monitoring what at an individual level i.e. how do you prevent people externally/internally utilises the value of this "interests disclosure"?
- How will you mitigate against "gaming" of news (as a secondary measure to reputation) whereby people collude to create noise in the blogosphere?

[Yaser Anwar]

Another Q I wanted to ask was: Would Monitor110 have only a select number of clients? Say 30-40. Why so? In my opinion, the early institutional clients won't have their alpha advantage diluted when more HFs get access to the data.

You might say, "Well the internet is always growing, so we can accommodate more and more clients." My response would be if Monitor110 really works like your team expects it to then IMO clients would be willing to pay a higher premium if they knew distribution would be limited.

Think of it like a specialized commodity that only very little can gain access to. Don't you think that would increase the value?

Who knows one day, the subscriptions could trade on a premium!

[Roger]

Bill, your characterization is incorrect so I'm letting you know. You are trying to stuff what the system is capable of doing and the people for whom it can create value in way, way too small of a box. I would say that the data could be applicable for all those whom you've mentioned. In fact, we have had several day-trading oriented shops (as well as more research-driven long/short, special situations and relative value shops) contact us about becoming beta clients.

That said, you are operating in a bit in an information vacuum having never seen the tool. In the absence of objective data, your skepticism is understandable.

[Bill aka NO DooDahs]

As I (mis?)understand it, M110 is a tool for traders to convert information into opportunity, and because of its nature, I would imagine it's primarily a tool for traders with mid-to-long holding periods (week or more). I doubt that the information generated is suitable to day trading, nor would the customer (as I (mis)understand you) be interested in short-term trading. If that's incorrect, let me know.

I never meant to say it would replace any other tool, but would be one additional tool out of many available. I'm skeptical about the ROI, but I'm usually skeptical by nature, so that's nothing new.

:-)

[Roger]

Gentlemen, sorry for my delayed response. So first:

Yaser: We do, in fact, tickerize the unstructured universe. This happens in real-time. We also give our clients the ability to modify the amount of data presented to them by varying thresholds of both reputation and relevance. The system does much more than that but this is a core part of the application. Concerning why not start a hedge fund given the power of the tool, my partner is a very successful technology entrepreneur and I am former Wall Street executive that ran a large trading business but was not a trader myself. Therefore, we made the decision to become an "arms dealer," to develop a tool that would have tremendous value in the hands of skilled analysts and traders. Data in a vacuum is pretty useless, but it is when this highly filtered data is presented in an intuitive way to these skilled practitioners that the value is created. Hopefully this answers your question.

Jordan: Monitor110 is powerful platform for harvesting, matching, filtering, analyzing, visualizing and displaying unstructured data from the internet. We are currently running the "free web" data across our platform. Clients will have the ability to run either proprietary or subscription feeds through our system as well. We are not developing bespoke branded content. We are helping our institutional investor clients discover valuable content out on the internet that is relevant to them from sources that meet their designated thresholds of reputability.

Bill: I get what you are saying but I'm not sure you understand the tool. Monitor110 isn't simply designed to be a whiz-bang trading application (though aspects of this are on our development roadmap). It is a powerful tool for discovery and analysis that can be used by research analysts and portfolio managers alike, as well as possessing sophisticated alerting tools for traders driven off of metadata and events happening out on the internet. Monitor110 is not specifically designed to replace technical analysis, nor is technical analysis remotely close to generating the value that can be extracted by a Monitor110 user.

Thanks, as always, for your comments and thanks also for the well-wishes. Right back at you.

[Bill aka NO DooDahs]

Yep, Seeking Alpha also got "mentioned" on that first-page-but-below-the-fold article in FT, just not by name. I get the FT in paper format via mail plus have internet access. The whiz-bang marketing gurus at FT actually made it cheaper for me to get paper and internet than it was to get either seperately. Go figure.

I wish y'all the best in Monitor110, but I wonder if it's any better (considering the expense) than simple fundamental and technical analysis. I consider myself an amateur technician at best and saw and wrote about bearish signs on GOOG in December. In terms of valuation, the signs were there much earlier.

http://stockcharts.com/h-sc/ui?s=GOOG&p=D&yr=1&mn=0&dy=0&id=p02660721804

http://www.marketthoughts.com/forum/viewtopic.php?p=3484#3484

http://www.marketthoughts.com/forum/viewtopic.php?t=1407

An options trader could have done a simple bet on volatility come earnings time to make some money. It was that obvious GOOG would move violently come earnings time.

So ... is data mining the blogosphere going to provide ROI? I'm skeptical at this point, but that doesn't mean I don't wish you well ...

[Jordan]

Congratulations on the launch! We're all eager to see what's next! Reuters and Bloomberg depend on "silo's" in their branded content, does Monitor have to?

[Yaser Anwar]

Congratulations!! Monitor110 looks like a real solid idea.

Deciphering the plethora of information online into tickerized content according to relevance. Did I get it right?

Another Q- Why do you invest in HFs? I only ask this because you ran a 130-person team of one of the biggest Wall St houses & have the amazing tool Monitor110, wouldn't you be better of investing on your own? (Maybe its just because you don't have time)

My only regret is Monitor110 isn't available to individual investors. With the examples provided at http://monitor110.com/monetize.php
sure seems one hell of a tool!

Best of luck with Monitor110!