Opening Cross: Predictive Text

max-bowie

If you've been reading the news, you'll know that a religious group has spent a great deal of money advertising its belief that the world will end on Saturday, May 21. And if you're reading this, then you'll know that didn't happen. But how do you determine whether someone's prediction made on their chosen set of data should be discounted or accepted?

After all, every prediction - whether it be a trade placed in response to a news story or economic event based on how the markets have reacted to similar news in the past, such as those currently being tested by Brazil's Itaú-Unibanco Asset Management (see story, page 1), or estimates of spending trends on market data and related services based on a survey of market participants, such as that carried out by Burton-Taylor International Consulting (see story, page 6) - is a theory until facts prove it right or wrong.

Of course, these types of predictions are more than just educated guesses. And one reason it would be such a shame were the world to end on May 21 would be that you'd never get to read about them and find out the challenges Itaú faces in finding sufficient local historical data to support news-based algorithmic trading, or how Burton-Taylor predicts that risk, commodities and regulation will be the drivers of data spend, and how China and India will be the big-growth consumers of data in the next two years.

You'd also miss next week's annual North American Financial Information Summit in New York, where industry practitioners meet to debate the issues affecting market data and reference data professionals. That would mean missing out on a keynote address from Lou Eccleston as he guides the newly-formed McGraw-Hill Financial through the process of separating from parent Standard & Poor's, as well as panels covering data administration, sourcing and negotiation issues, regulation and the search for a universal legal entity identifier, enterprise data management and governance, using data to build optimized trading strategies, managing content ranging from low-latency data to corporate actions, ensuring accuracy of evaluated prices and counterparty data, and the impact of new initiatives and business models at exchanges in a rapidly-fragmenting landscape.

In addition, you'd never find out the winners of this year's Inside Market Data and Inside Reference Data Awards, which will be announced following the conference - though who's going to be crowing about their win in the afterlife (And the winner for best events calendar is... The Mayans)? - or who will be this year's inductee into the Inside Market Data Hall of Fame.

In the event of Armageddon, forward prices wouldn't matter anymore, such as those being launched by energy commodities broker OTC Global Holdings (see story, page 5). Nor would the data from Standard & Poor's Credit Health Panel, which last week rolled out a series of enhancements for users to customize the information they receive from the platform (see story, page 11).

And as for the quest for lower latency, forget it. Because if the world really is going to end on Saturday, then we needn't worry about any of this... or about doing laundry this evening. But let's assume that this radical data analysis is wrong - that the doom-mongers' predictions are an outlier, if you will - then we can discount it as a curious anomaly and work on the consensus opinion that Sunday will find the Earth still spinning, and Monday will find the markets open as always, data continuing to surge over the world's financial infrastructure, and all of us back at work as usual. Of course - unlike if I made a bad trade based on erroneous data - even if I'm wrong, who's going to be around to call me on it?

I hope - really hope - to see you all at the conference on Tuesday.

  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have a WatersTechnology account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: